Assessing Etiologic Heterogeneity for Multinomial Outcome with Two-Phase Outcome-Dependent Sampling Design.

Etiologic heterogeneity occurs when distinct sets of events or exposures give rise to different subtypes of disease. Inference about subtype-specific exposure effects from two-phase outcome-dependent sampling data requires adjustment for both confounding and the sampling design. Common approaches to inference for these effects do not necessarily appropriately adjust for these sources of bias, or allow for formal comparisons of effects across different subtypes. Herein, using inverse probability weighting (IPW) to fit a multinomial model is shown to yield valid inference with this sampling design for subtype-specific exposure effects and contrasts thereof. The IPW approach is compared to common regression-based methods for assessing exposure effect heterogeneity using simulations. The methods are applied to estimate subtype-specific effects of various exposures on breast cancer risk in the Carolina Breast Cancer Study.

Detecting responsible nodes in differential Bayesian networks.

To study the roles that different nodes play in differentiating Bayesian networks under two states, such as control versus disease, we formulate two node-specific scores to facilitate such assessment. The first score is motivated by the prediction invariance property of a causal model. The second score results from modifying an existing score constructed for differential analysis of undirected networks. We develop strategies based on these scores to identify nodes responsible for topological differences between two Bayesian networks. Synthetic data and real-life data from designed experiments are used to demonstrate the efficacy of the proposed methods in detecting responsible nodes.

Application of causal inference methods in individual-participant data meta-analyses in medicine: addressing data handling and reporting gaps with new proposed reporting guidelines.

Observational data provide invaluable real-world information in medicine, but certain methodological considerations are required to derive causal estimates. In this systematic review, we evaluated the methodology and reporting quality of individual-level patient data meta-analyses (IPD-MAs) conducted with non-randomized exposures, published in 2009, 2014, and 2019 that sought to estimate a causal relationship in medicine. We screened over 16,000 titles and abstracts, reviewed 45 full-text articles out of the 167 deemed potentially eligible, and included 29 into the analysis. Unfortunately, we found that causal methodologies were rarely implemented, and reporting was generally poor across studies. Specifically, only three of the 29 articles used quasi-experimental methods, and no study used G-methods to adjust for time-varying confounding. To address these issues, we propose stronger collaborations between physicians and methodologists to ensure that causal methodologies are properly implemented in IPD-MAs. In addition, we put forward a suggested checklist of reporting guidelines for IPD-MAs that utilize causal methods. This checklist could improve reporting thereby potentially enhancing the quality and trustworthiness of IPD-MAs, which can be considered one of the most valuable sources of evidence for health policy.

Development and evaluation of regression tree models for predicting in-hospital mortality of a national registry of COVID-19 patients over six pandemic surges.

BACKGROUND: Objective prognostic information is essential for good clinical decision making. In case of unknown diseases, scarcity of evidence and limited tacit knowledge prevent obtaining this information. Prediction models can be useful, but need to be not only evaluated on how well they predict, but also how stable these models are under fast changing circumstances with respect to development of the disease and the corresponding clinical response. This study aims to provide interpretable and actionable insights, particularly for clinicians. We developed and evaluated two regression tree predictive models for in-hospital mortality of COVID-19 patient at admission and 24 hours (24 h) after admission, using a national registry. We performed a retrospective analysis of observational routinely collected data. METHODS: Two regression tree models were developed for admission and 24 h after admission. The complexity of the trees was managed via cross validation to prevent overfitting. The predictive ability of the model was assessed via bootstrapping using the Area under the Receiver-Operating-Characteristic curve, Brier score and calibration curves. The tree models were assessed on the stability of their probabilities and predictive ability, on the selected variables, and compared to a full-fledged logistic regression model that uses variable selection and variable transformations using splines. Participants included COVID-19 patients from all ICUs participating in the Dutch National Intensive Care Evaluation (NICE) registry, who were admitted at the ICU between February 27, 2020, and November 23, 2021. From the NICE registry, we included concerned demographic data, minimum and maximum values of physiological data in the first 24 h of ICU admission and diagnoses (reason for admission as well as comorbidities) for model development. The main outcome measure was in-hospital mortality. We additionally analysed the Length-of-Stay (LoS) per patient subgroup per survival status. RESULTS: A total of 13,369 confirmed COVID-19 patients from 70 ICUs were included (with mortality rate of 28%). The optimism-corrected AUROC of the admission tree (with seven paths) was 0.72 (95% CI: 0.71-0.74) and of the 24 h tree (with 11 paths) was 0.74 (0.74-0.77). Both regression trees yielded good calibration and variable selection for both trees was stable. Patient subgroups comprising the tree paths had comparable survival probabilities as the full-fledged logistic regression model, survival probabilities were stable over six COVID-19 surges, and subgroups were shown to have added predictive value over the individual patient variables. CONCLUSIONS: We developed and evaluated regression trees, which operate at par with a carefully crafted logistic regression model. The trees consist of homogenous subgroups of patients that are described by simple interpretable constraints on patient characteristics thereby facilitating shared decision-making.

Association of hospital-based substance use supports on emergency department revisits: a retrospective cohort study in Sudbury, Canada from 2018 to 2022.

BACKGROUND: This study compares emergency department (ED) revisits for patients receiving hospital-based substance-use support compared to those who did not receive specialized addiction services at Health Sciences North in Sudbury, Ontario, Canada. METHODS: The study is a retrospective observational study using administrative data from all patients presenting with substance use disorder (SUD) at Health Sciences North from January 1, 2018, and August 31, 2022 with ICD-10 codes from the Discharge Abstract Database (DAD) and the National Ambulatory Care Database (NACRS). There were two interventions under study: addiction medicine consult services (AMCS group), and specialized addiction medicine unit (AMU group). The AMCS is a consult service offered for patients in the ED and those who are admitted to the hospital. The AMU is a specialized inpatient medical unit designed to offer addiction support to stabilize patients that operates under a harm-reduction philosophy. The primary outcome was all cause ED revisit within 30 days of the index ED or hospital visit. The secondary outcome was all observed ED revisits in the study period. Kaplan-Meier curves were used to measure the proportion of 30-day revisits by exposure group. Odds ratios and Hazard Ratios were calculated using logistic regression models with random effects and Cox-proportional hazard model respectively. RESULTS: A total of 5,367 patients with 10,871 ED index visits, and 2,127 revisits between 2018 and 2022 are included in the study. 45% (2,340/5,367) of patient were not admitted to hospital. 30-day revisits were less likely among the intervention group: Addiction Medicine Consult Services (AMCS) in the ED significantly reduced the odds of revisits (OR 0.53, 95% CI 0.39-0.71, p < 0.01) and first revisits (OR 0.42, 95% CI 0.33-0.53, p < 0.01). The AMU group was associated with lower revisits odds (OR 0.80, 95% CI 0.66-0.98, p = 0.03). For every additional year of age, the odds of revisits slightly decreased (OR 0.99, 95% CI 0.98-1.00, p = 0.01) and males were found to have an increased risk compared to females (OR 1.50, 95% CI 1.35-1.67, p < 0.01). INTERPRETATION: We observe statistically significant differences in ED revisits for patients receiving hospital-based substance-use support at Health Sciences North. Hospital-based substance-use supports could be applied to other hospitals to reduce 30-day revisits.

Current Challenges With the Use of Test-Negative Designs for Modeling COVID-19 Vaccination and Outcomes.

The widespread testing for severe acute respiratory syndrome coronavirus 2 infection has facilitated the use of test-negative designs (TNDs) for modeling coronavirus disease 2019 (COVID-19) vaccination and outcomes. Despite the comprehensive literature on TND, the use of TND in COVID-19 studies is relatively new and calls for robust design and analysis to adapt to a rapidly changing and dynamically evolving pandemic and to account for changes in testing and reporting practices. In this commentary, we aim to draw the attention of researchers to COVID-specific challenges in using TND as we are analyzing data amassed over more than two years of the pandemic. We first review when and why TND works and general challenges in TND studies presented in the literature. We then discuss COVID-specific challenges which have not received adequate acknowledgment but may add to the risk of invalid conclusions in TND studies of COVID-19.

The effects of hypothetical behavioral interventions on the 13-year incidence of overweight/obesity in children and adolescents.

BACKGROUND: In view of the high burden of childhood overweight/obesity (OW/OB), it is important to identify targets for interventions that may have the greatest effects on preventing OW/OB in early life. Using methods of causal inference, we studied the effects of sustained behavioral interventions on the long-term risk of developing OW/OB based on a large European cohort. METHODS: Our sample comprised 10 877 children aged 2 to < 10 years at baseline who participated in the well-phenotyped IDEFICS/I.Family cohort. Children were followed from 2007/08 to 2020/21. Applying the parametric g-formula, the 13-year risk of developing OW/OB was estimated under various sustained hypothetical interventions on physical activity, screen time, dietary intake and sleep duration. Interventions imposing adherence to recommendations (e.g. maximum 2 h/day screen time) as well as interventions 'shifting' the behavior by a specified amount (e.g. decreasing screen time by 30 min/day) were compared to 'no intervention' (i.e. maintaining the usual or so-called natural behavior). Separately, the effectiveness of these interventions in vulnerable groups was assessed. RESULTS: The 13-year risk of developing OW/OB was 30.7% under no intervention and 25.4% when multiple interventions were imposed jointly. Meeting screen time and moderate-to-vigorous physical activity (MVPA) recommendations were found to be most effective, reducing the incidence of OW/OB by -2.2 [-4.4;-0.7] and -2.1 [-3.7;-0.8] percentage points (risk difference [95% confidence interval]), respectively. Meeting sleep recommendations (-0.6 [-1.1;-0.3]) had a similar effect as increasing sleep duration by 30 min/day (-0.6 [-0.9;-0.3]). The most effective intervention in children of parents with low/medium educational level was being member in a sports club; for children of mothers with OW/OB, meeting screen time recommendations and membership in a sports club had the largest effects. CONCLUSIONS: While the effects of single behavioral interventions sustained over 13 years were rather small, a joint intervention on multiple behaviors resulted in a relative reduction of the 13-year OW/OB risk by between 10 to 26%. Individually, meeting MVPA and screen time recommendations were most effective. Nevertheless, even under the joint intervention the absolute OW/OB risk remained at a high level of 25.4% suggesting that further strategies to better prevent OW/OB are required.

Implementation of the trial emulation approach in medical research: a scoping review.

BACKGROUND: When conducting randomised controlled trials is impractical, an alternative is to carry out an observational study. However, making valid causal inferences from observational data is challenging because of the risk of several statistical biases. In 2016 Hernán and Robins put forward the 'target trial framework' as a guide to best design and analyse observational studies whilst preventing the most common biases. This framework consists of (1) clearly defining a causal question about an intervention, (2) specifying the protocol of the hypothetical trial, and (3) explaining how the observational data will be used to emulate it. METHODS: The aim of this scoping review was to identify and review all explicit attempts of trial emulation studies across all medical fields. Embase, Medline and Web of Science were searched for trial emulation studies published in English from database inception to February 25, 2021. The following information was extracted from studies that were deemed eligible for review: the subject area, the type of observational data that they leveraged, and the statistical methods they used to address the following biases: (A) confounding bias, (B) immortal time bias, and (C) selection bias. RESULTS: The search resulted in 617 studies, 38 of which we deemed eligible for review. Of those 38 studies, most focused on cardiology, infectious diseases or oncology and the majority used electronic health records/electronic medical records data and cohort studies data. Different statistical methods were used to address confounding at baseline and selection bias, predominantly conditioning on the confounders (N = 18/49, 37%) and inverse probability of censoring weighting (N = 7/20, 35%) respectively. Different approaches were used to address immortal time bias, assigning individuals to treatment strategies at start of follow-up based on their data available at that specific time (N = 21, 55%), using the sequential trial emulations approach (N = 11, 29%) or the cloning approach (N = 6, 16%). CONCLUSION: Different methods can be leveraged to address (A) confounding bias, (B) immortal time bias, and (C) selection bias. When working with observational data, and if possible, the 'target trial' framework should be used as it provides a structured conceptual approach to observational research.

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.

Quasi-rerandomization for observational studies.

BACKGROUND: In the causal analysis of observational studies, covariates should be carefully balanced to approximate a randomized experiment. Numerous covariate balancing methods have been proposed for this purpose. However, it is often unclear what type of randomized experiments the balancing approaches aim to approximate; and this may cause ambiguity and hamper the synthesis of balancing characteristics within randomized experiments. METHODS: Randomized experiments based on rerandomization, known for significant improvement on covariate balance, have recently gained attention in the literature, but no attempt has been made to integrate this scheme into observational studies for improving covariate balance. Motivated by the above concerns, we propose quasi-rerandomization, a novel reweighting method, where observational covariates are rerandomized to be the anchor for reweighting such that the balanced covariates obtained from rerandomization can be reconstructed by the weighted data. RESULTS: Through extensive numerical studies, not only does our approach demonstrate similar covariate balance and comparable estimation precision of treatment effect to rerandomization in many situations, but it also exhibits advantages over other balancing techniques in inferring the treatment effect. CONCLUSION: Our quasi-rerandomization method can approximate the rerandomized experiments well in terms of improving the covariate balance and the precision of treatment effect estimation. Furthermore, our approach shows competitive performance compared with other weighting and matching methods. The codes for the numerical studies are available at https://github.com/BobZhangHT/QReR .

Learning end-to-end patient representations through self-supervised covariate balancing for causal treatment effect estimation.

A causal effect can be defined as a comparison of outcomes that result from two or more alternative actions, with only one of the action-outcome pairs actually being observed. In healthcare, the gold standard for causal effect measurements is randomized controlled trials (RCTs), in which a target population is explicitly defined and each study sample is randomly assigned to either the treatment or control cohorts. The great potential to derive actionable insights from causal relationships has led to a growing body of machine-learning research applying causal effect estimators to observational data in the fields of healthcare, education, and economics. The primary difference between causal effect studies utilizing observational data and RCTs is that for observational data, the study occurs after the treatment, and therefore we do not have control over the treatment assignment mechanism. This can lead to massive differences in covariate distributions between control and treatment samples, making a comparison of causal effects confounded and unreliable. Classical approaches have sought to solve this problem piecemeal, first by predicting treatment assignment and then treatment effect separately. Recent work extended part of these approaches to a new family of representation-learning algorithms, showing that the upper bound of the expected treatment effect estimation error is determined by two factors: the outcome generalization-error of the representation and the distance between the treated and control distributions induced by the representation. To achieve minimal dissimilarity in learning such distributions, in this work we propose a specific auto-balancing, self-supervised objective. Experiments on real and benchmark datasets revealed that our approach consistently produced less biased estimates than previously published state-of-the-art methods. We demonstrate that the reduction in error can be directly attributed to the ability to learn representations that explicitly reduce such dissimilarity; further, in case of violations of the positivity assumption (frequent in observational data), we show our approach performs significantly better than the previous state of the art. Thus, by learning representations that induce similar distributions of the treated and control cohorts, we present evidence to support the error bound dissimilarity hypothesis as well as providing a new state-of-the-art model for causal effect estimation.

Comparing broad and narrow phenotype algorithms: differences in performance characteristics and immortal time incurred.

Introduction: When developing phenotype algorithms for observational research, there is usually a trade-off between definitions that are sensitive or specific. The objective of this study was to estimate the performance characteristics of phenotype algorithms designed for increasing specificity and to estimate the immortal time associated with each algorithm. Materials and methods: We examined algorithms for 11 chronic health conditions. The analyses were from data from five databases. For each health condition, we created five algorithms to examine performance (sensitivity and positive predictive value (PPV)) differences: one broad algorithm using a single code for the health condition and four narrow algorithms where a second diagnosis code was required 1-30 days, 1-90 days, 1-365 days, or 1- all days in a subject's continuous observation period after the first code. We also examined the proportion of immortal time relative to time-at-risk (TAR) for four outcomes. The TAR's were: 0-30 days after the first condition occurrence (the index date), 0-90 days post-index, 0-365 days post-index, and 0-1,095 days post-index. Performance of algorithms for chronic health conditions was estimated using PheValuator (V2.1.4) from the OHDSI toolstack. Immortal time was calculated as the time from the index date until the first of the following: 1) the outcome; 2) the end of the outcome TAR; 3) the occurrence of the second code for the chronic health condition. Results: In the first analysis, the narrow phenotype algorithms, i.e., those requiring a second condition code, produced higher estimates for PPV and lower estimates for sensitivity compared to the single code algorithm. In all conditions, increasing the time to the required second code increased the sensitivity of the algorithm. In the second analysis, the amount of immortal time increased as the window used to identify the second diagnosis code increased. The proportion of TAR that was immortal was highest in the 30 days TAR analyses compared to the 1,095 days TAR analyses. Conclusion: Attempting to increase the specificity of a health condition algorithm by adding a second code is a potentially valid approach to increase specificity, albeit at the cost of incurring immortal time.

Prescription drug monitoring programs use mandates and prescription stimulant and depressant quantities.

BACKGROUND: While the mandate to check patients' prescription history in Prescription Drug Monitoring Program (PDMP) database before prescribing/dispensing controlled drugs has been shown to be an important tool to curb opioid abuse, less is known about whether the mandate can reduce the misuse of other commonly abused prescription drugs. We examined whether PDMP use mandates were associated with changes in prescription stimulant and depressant quantities. METHODS: Using data from Automated Reports and Consolidate Ordering System (ARCOS), we employed difference-in-differences design to estimate the association between PDMP use mandates and prescription stimulant and depressant quantities in 50 U.S. states and the District of Columbia from 2006 to 2020. Limited PDMP use mandate was specific only to opioids or benzodiazepines. Expansive PDMP use mandate was non-specific to opioid or benzodiazepine and required prescribers/dispensers to check PDMP when prescribing/dispensing targeted controlled substances in Schedule II-V. The main outcomes were population-adjusted prescription stimulant (amphetamine, methylphenidate, lisdexamfetamine) and depressant (amobarbital, butalbital, pentobarbital, secobarbital) quantities in grams. RESULTS: There was no evidence that limited PDMP use mandate was associated with a reduction in the prescription stimulant and depressant quantities. However, expansive PDMP use mandate that was non-specific to opioid or benzodiazepine and required prescribers/dispensers to check PDMP when prescribing/dispensing targeted controlled substances in Schedule II-V was associated with 6.2% (95% CI: -10.06%, -2.08%) decline in prescription amphetamine quantity. CONCLUSION: Expansive PDMP use mandate was associated with a decline in prescription amphetamine quantity. Limited PDMP use mandate did not appear to change prescription stimulant and depressant quantities.

Dealing with missing data in laboratory test results used as a baseline covariate: results of multi-hospital cohort studies utilizing a database system contributing to MID-NET® in Japan.

BACKGROUND: To evaluate missing data methods applied to laboratory test results used for confounding adjustment, utilizing data from 10 MID-NET®-collaborative hospitals. METHODS: Using two scenarios, five methods dealing with missing laboratory test results were applied, including three missing data methods (single regression imputation (SRI), multiple imputation (MI), and inverse probability weighted (IPW) method). We compared the point estimates of adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) between the five methods. Hospital variability in missing data was considered using the hospital-specific approach and overall approach. Confounding adjustment methods were propensity score (PS) weighting, PS matching, and regression adjustment. RESULTS: In Scenario 1, the risk of diabetes due to second-generation antipsychotics was compared with that due to first-generation antipsychotics. The aHR adjusted by PS weighting using SRI, MI, and IPW by the hospital-specific-approach was 0.61 [95%CI, 0.39-0.96], 0.63 [95%CI, 0.42-0.93], and 0.76 [95%CI, 0.46-1.25], respectively. In Scenario 2, the risk of liver injuries due to rosuvastatin was compared with that due to atorvastatin. Although PS matching largely contributed to differences in aHRs between methods, PS weighting provided no substantial difference in point estimates of aHRs between SRI and MI, similar to Scenario 1. The results of SRI and MI in both scenarios showed no considerable changes, even upon changing the approaches considering hospital variations. CONCLUSIONS: SRI and MI provide similar point estimates of aHR. Two approaches considering hospital variations did not markedly affect the results. Adjustment by PS matching should be used carefully.

Deep learning and differential equations for modeling changes in individual-level latent dynamics between observation periods.

When modeling longitudinal biomedical data, often dimensionality reduction as well as dynamic modeling in the resulting latent representation is needed. This can be achieved by artificial neural networks for dimension reduction and differential equations for dynamic modeling of individual-level trajectories. However, such approaches so far assume that parameters of individual-level dynamics are constant throughout the observation period. Motivated by an application from psychological resilience research, we propose an extension where different sets of differential equation parameters are allowed for observation subperiods. Still, estimation for intra-individual subperiods is coupled for being able to fit the model also with a relatively small dataset. We subsequently derive prediction targets from individual dynamic models of resilience in the application. These serve as outcomes for predicting resilience from characteristics of individuals, measured at baseline and a follow-up time point, and selecting a small set of important predictors. Our approach is seen to successfully identify individual-level parameters of dynamic models that allow to stably select predictors, that is, resilience factors. Furthermore, we can identify those characteristics of individuals that are the most promising for updates at follow-up, which might inform future study design. This underlines the usefulness of our proposed deep dynamic modeling approach with changes in parameters between observation subperiods.

On Variance of the Treatment Effect in the Treated When Estimated by Inverse Probability Weighting.

In the analysis of observational studies, inverse probability weighting (IPW) is commonly used to consistently estimate the average treatment effect (ATE) or the average treatment effect in the treated (ATT). The variance of the IPW ATE estimator is often estimated by assuming that the weights are known and then using the so-called "robust" (Huber-White) sandwich estimator, which results in conservative standard errors (SEs). Here we show that using such an approach when estimating the variance of the IPW ATT estimator does not necessarily result in conservative SE estimates. That is, assuming the weights are known, the robust sandwich estimator may be either conservative or anticonservative. Thus, confidence intervals for the ATT using the robust SE estimate will not be valid, in general. Instead, stacked estimating equations which account for the weight estimation can be used to compute a consistent, closed-form variance estimator for the IPW ATT estimator. The 2 variance estimators are compared via simulation studies and in a data analysis of the association between smoking and gene expression.

Long-term treatment patterns and survival in metastatic breast cancer by intrinsic subtypes - an observational cohort study in Sweden.

BACKGROUND: Longitudinal, real-world data on the management of metastatic breast cancer is increasingly relevant to understand breast cancer care in routine clinical practice. Yet such data are scarce, particularly beyond second- and third-line treatment strategies. This study, therefore, examined both the long-term treatment patterns and overall survival (OS) in a regional Swedish cohort of female patients with metastatic breast cancer stratified by subtype in routine clinical practice during a recent eight-year period and correlation to current treatment guidelines. METHODS: Consecutive female patients with metastatic breast cancer clinically managed at Uppsala University Hospital, Uppsala, Sweden, during 2009-2016 and followed until the end of September, 2017 (n = 370) were included and, where possible, classified as having one of five, intrinsic subtypes: Luminal A; Luminal B; human epidermal growth factor receptor 2-positive (HER2+)/ estrogen receptor-positive (ER+); HER2+/estrogen receptor-negative (ER-); or triple negative breast cancer (TNBC). Treatment patterns and OS were estimated by subtype using time-to-event methods. RESULTS: A total of 352/370 patients with metastatic breast cancer (mean age 67.6 years) could be subtyped: 118 (34%) were Luminal A, 119 (34%) Luminal B, 31 (8%) HER2+/ER-, 38 (11%) HER2+/Luminal, and 46 (13%) TNBC. The median number of metastatic treatment lines was 3. Most patients were on active treatment during follow-up (80% of the observation period), except for patients with TNBC who were on treatment for 60% of the observation time. Overall, 67% of patients died whilst on treatment. Among all patients (n = 370), median OS was 32.5 months (95% CI = 28.2-35.7). The 5-year survival rate was highest for HER2+/Luminal (46%) patients, followed by Luminal B (29%), Luminal A (28%), HER2+/ER- (21%), and TNBC (7%). Increasing age and number of metastatic sites also predicted worse survival. CONCLUSIONS: Metastatic breast cancer patients in Sweden, irrespective of subtype, generally receive active treatment until time of death. Survival varies considerably across subtypes and is also associated with patient characteristics. Regardless of differences in treatment patterns for Luminal A and B patients, long-term OS was the same.

Achieving diabetes treatment targets in people with registered mental illness is similar or improved compared with those without: Analyses of linked observational datasets.

AIMS: To determine the association between registered mental illness and type 2 diabetes mellitus treatment targets, while taking into account the effects of health expenditure and social determinants of health. METHODS: This observational cross-sectional study was based on routine primary care data, linked to socio-economic and medical claims data. The main outcomes, analysed by multivariate logistic regression, were achieving primary care guideline treatment targets for HbA1c , systolic blood pressure (SBP) and LDL-cholesterol in 2017. We examined the association with diagnosed mental illness registered by the general practitioner (GP) or treated via specialist' mental healthcare between 2016 and 2018, adjusting for, medication use, body mass index, co-morbidity, smoking, and additionally examining effect-modification of healthcare expenditures, migration status, income and demographics. RESULTS: Overall (N = 2862), 64.0% of participants achieved their treatment targets for HbA1c , 65.1% for SBP and 53.0% for LDL-cholesterol. Adjusted for migrant background, income and care expenditures, individuals <65 years of age with mental illness achieved their HbA1c treatment target more often than those without (OR (95% CI)): treatment by GP: 1.46 (1.01, 2.11), specialist care: 1.61 (1.11, 2.34), as did men with mental illness for SBP: GP OR 1.61 (1.09, 2.40), specialist care OR 1.59 (1.09, 2.45). LDL-cholesterol target was not associated with mental illness. A migrant background or low income lowered the likelihood of reaching HbA1c targets. CONCLUSIONS: People with registered mental illness appear comparable or better able to achieve diabetes treatment targets than those without. Achieving HbA1c targets is influenced by social disadvantage.

Instrumental variable analysis to estimate treatment effects: a simulation study showing potential benefits of conditioning on hospital.

BACKGROUND: Instrumental variable (IV) analysis holds the potential to estimate treatment effects from observational data. IV analysis potentially circumvents unmeasured confounding but makes a number of assumptions, such as that the IV shares no common cause with the outcome. When using treatment preference as an instrument, a common cause, such as a preference regarding related treatments, may exist. We aimed to explore the validity and precision of a variant of IV analysis where we additionally adjust for the provider: adjusted IV analysis. METHODS: A treatment effect on an ordinal outcome was simulated (beta - 0.5 in logistic regression) for 15.000 patients, based on a large data set (the IMPACT data, n = 8799) using different scenarios including measured and unmeasured confounders, and a common cause of IV and outcome. We compared estimated treatment effects with patient-level adjustment for confounders, IV with treatment preference as the instrument, and adjusted IV, with hospital added as a fixed effect in the regression models. RESULTS: The use of patient-level adjustment resulted in biased estimates for all the analyses that included unmeasured confounders, IV analysis was less confounded, but also less reliable. With correlation between treatment preference and hospital characteristics (a common cause) estimates were skewed for regular IV analysis, but not for adjusted IV analysis. CONCLUSION: When using IV analysis for comparing hospitals, some limitations of regular IV analysis can be overcome by adjusting for a common cause. TRIAL REGISTRATION: We do not report the results of a health care intervention.

Empirical assessment of alternative methods for identifying seasonality in observational healthcare data.

BACKGROUND: Seasonality classification is a well-known and important part of time series analysis. Understanding the seasonality of a biological event can contribute to an improved understanding of its causes and help guide appropriate responses. Observational data, however, are not comprised of biological events, but timestamped diagnosis codes the combination of which (along with additional requirements) are used as proxies for biological events. As there exist different methods for determining the seasonality of a time series, it is necessary to know if these methods exhibit concordance. In this study we seek to determine the concordance of these methods by applying them to time series derived from diagnosis codes in observational data residing in databases that vary in size, type, and provenance. METHODS: We compared 8 methods for determining the seasonality of a time series at three levels of significance (0.01, 0.05, and 0.1), against 10 observational health databases. We evaluated 61,467 time series at each level of significance, totaling 184,401 evaluations. RESULTS: Across all databases and levels of significance, concordance ranged from 20.2 to 40.2%. Across all databases and levels of significance, the proportion of time series classified seasonal ranged from 4.9 to 88.3%. For each database and level of significance, we computed the difference between the maximum and minimum proportion of time series classified seasonal by all methods. The median within-database difference was 54.8, 34.7, and 39.8%, for p < 0.01, 0.05, and 0.1, respectively. CONCLUSION: Methods of binary seasonality classification when applied to time series derived from diagnosis codes in observational health data produce inconsistent results. The methods exhibit considerable discord within all databases, implying that the discord is a result of the difference between the methods themselves and not due to the choice of database. The results indicate that researchers relying on automated methods to assess the seasonality of time series derived from diagnosis codes in observational data should be aware that the methods are not interchangeable and thus the choice of method can affect the generalizability of their work. Seasonality determination is highly dependent on the method chosen.