Explore the vast range of titles available.

The randomised trial is the preferred study design for evaluating the effectiveness and safety of interventions. Yet such trials can be prohibitively expensive, unethical, or take too long. When it is not possible to carry out a randomised trial, observational data can be used to answer similar questions. Here, we describe the process of using observational data to emulate a target trial, which applies the study design principles of randomised trials to observational studies that aim to estimate the causal effect of an intervention. The target trial provides a formal framework to help avoid self-inflicted biases common to observational studies.

Benchmarking an observational analysis against a randomized trial can increase confidence in the use of observational data to complement inferences made in trials. Until now, few examples of benchmarking have been within oncology. However, benchmarking trials of a cancer treatment poses a unique set of challenges, such as defining composite outcomes like disease-free survival. We designed a target trial with a protocol as similar as possible to the B-31 and N9831 randomized trials, which estimated the effect of adjuvant trastuzumab plus chemotherapy compared with chemotherapy alone in individuals with early human epidermal growth factor receptor 2-positive breast cancer. We then carried out an observational analysis by emulating the target trial using routinely collected data from Swedish registries to understand if we can estimate a similar effect of trastuzumab as the trial. The primary endpoint was the composite of disease-free survival consisting of the earliest of (1) local or regional recurrences, (2) distant recurrences, (3) contralateral breast cancer, (4) other second primary cancer, or (5) death from any cause. Individuals who had data compatible with both treatment strategies at baseline were cloned and one copy was assigned to each arm. We applied inverse probability weights to adjust for baseline and time-varying confounding (e.g., age and hematological events like neutropenia). Our observational analysis included 1,578 women, with a median age of 59 years, and who were diagnosed between 2008 and 2015. We estimated a similar effect after five years of follow-up (RR: 0.54, 95% CI [0.44, 0.67]) for the composite endpoint of disease-free survival as the two jointly analyzed B-31 and N9831 trials (HR: 0.48, 95% CI [0.39, 0.59]). While the comparability of results increases confidence in our estimates, there remains a risk of residual and unmeasured confounding, as is the case with all observational analyses. We successfully benchmarked an observational analysis against the B-31 and N9831 trials. By aligning protocols and using appropriate methodological approaches, we show that observational data can be used to estimate similar results as randomized trials of cancer treatments, like trastuzumab. This opens the door to using observational data to complement results from randomized trials of cancer treatments which can provide quick, cheap, and robust evidence to support decision-making where trials leave evidence gaps.

ObjectivePhase II trials suggest glucagon-like peptide-1 receptor (GLP1) agonists resolve metabolic dysfunction-associated steatohepatitis but do not affect fibrosis regression. We aimed to determine the long-term causal effect of GLP1 agonists on the risk of major adverse liver outcomes (MALO) in patients with any chronic liver disease and type 2 diabetes.DesignWe used observational data from Swedish healthcare registers 2010–2020 to emulate a target trial of GLP1 agonists in eligible patients with chronic liver disease and type 2 diabetes. We used an inverse-probability weighted marginal structural model to compare parametric estimates of 10-year MALO risk (decompensated cirrhosis, hepatocellular carcinoma, liver transplantation or MALO-related death) in initiators of GLP1 agonists with non-initiators. We randomly sampled 5% of the non-initiators to increase computational efficiency.ResultsGLP1 agonist initiators had a 10-year risk of MALO at 13.3% (42/1026) vs 14.6% in non-initiators (1079/15 633) in intention-to-treat analysis (risk ratio (RR)=0.91, 95% CI=0.50 to 1.32). The corresponding 10-year per-protocol risk estimates were 7.4% (22/1026) and 14.4% (1079/15 633), respectively (RR=0.51, 95% CI=0.14 to 0.88). The per-protocol risk estimates at 6 years were 5.4% (21/1026) vs 9.0% (933/15 633) (RR=0.60, 95% CI=0.29 to 0.90) and at 8 years 7.2% (22/1026) vs 11.7% (1036/15 633) (RR=0.61, 95% CI=0.21 to 1.01).ConclusionIn patients with chronic liver disease and type 2 diabetes who adhered to therapy over time, GLP1 agonists may result in lower risk of MALO. This suggests that GLP1 agonists are promising agents to reduce risk of chronic liver disease progression in patients with concurrent type 2 diabetes, although this needs to be corroborated in randomised trials.

The past few decades have seen substantial improvements in cancer survival, but concerns exist about long-term cardiovascular disease risk in survivors. Evidence is scarce on the risks of specific cardiovascular diseases in survivors of a wide range of cancers to inform prevention and management. In this study, we used large-scale electronic health records data from multiple linked UK databases to address these evidence gaps. For this population-based cohort study, we used linked primary care, hospital, and cancer registry data from the UK Clinical Practice Research Datalink to identify cohorts of survivors of the 20 most common cancers who were 18 years or older and alive 12 months after diagnosis and controls without history of cancer, matched for age, sex, and general practice. We compared risks for a range of cardiovascular disease outcomes using crude and adjusted Cox models. We fitted interactions to investigate effect modification, and flexible parametric survival models to estimate absolute excess risks over time. Between Jan 1, 1990, and Dec 31, 2015, 126 120 individuals with a diagnosis of a cancer of interest still being followed up at least 1 year later were identified and matched to 630 144 controls. After exclusions, 108 215 cancer survivors and 523 541 controls were included in the main analyses. Venous thromboembolism risk was elevated in survivors of 18 of 20 site-specific cancers compared with that of controls; adjusted hazard ratios (HRs) ranged from 1·72 (95% CI 1·57–1·89) in patients after prostate cancer to 9·72 (5·50–17·18) after pancreatic cancer. HRs decreased over time, but remained elevated more than 5 years after diagnosis. We observed increased risks of heart failure or cardiomyopathy in patients after ten of 20 cancers, including haematological (adjusted HR 1·94, 1·66–2·25, with non-Hodgkin lymphoma; 1·77, 1·50–2·09, with leukaemia; and 3·29, 2·59–4·18, with multiple myeloma), oesophageal (1·96, 1·46–2·64), lung (1·82, 1·52–2·17) kidney (1·73, 1·38–2·17) and ovarian (1·59, 1·19–2·12). Elevated risks of arrhythmia, pericarditis, coronary artery disease, stroke, and valvular heart disease were also observed for multiple cancers, including haematological malignancies. HRs for heart failure or cardiomyopathy and venous thromboembolism were greater in patients without previous cardiovascular disease and in younger patients. However, absolute excess risks were generally greater with increasing age. Increased risks of these outcomes seemed most pronounced in patients who had received chemotherapy. Survivors of most site-specific cancers had increased medium-term to long-term risk for one or more cardiovascular diseases compared with that for the general population, with substantial variations between cancer sites. Wellcome Trust and Royal Society.

There is concern regarding the impact of selective serotonin reuptake inhibitors (SSRIs) on suicidal behaviour. Using the target trial framework, we investigated the effect on suicidal behaviour of SSRI treatment following a depression diagnosis. We identified 162,267 individuals receiving a depression diagnosis aged 6–59 years during 2006–2018 in Stockholm County, Sweden, after at least 1 year without antidepressant dispensation. Individuals who initiated an SSRI within 28 days of the diagnosis were assigned as SSRI initiators, others as non-initiators. Intention-to-treat and per-protocol effects were estimated; for the latter, individuals were censored when they ceased adhering to their assigned treatment strategy. We applied inverse probability weighting (IPW) to account for baseline confounding in the intention-to-treat analysis, and additionally for treatment non-adherence and time-varying confounding in the per-protocol analysis. The suicidal behaviour risk difference (RD), and risk ratio (RR) between SSRI initiators and non-initiators were estimated at 12 weeks. In the overall cohort, we found an increased risk of suicidal behaviour among SSRI initiators (intention-to-treat RR = 1.50, 95% CI = 1.25, 1.80; per-protocol RR = 1.69, 95% CI = 1.20, 2.36). In age strata, we only found evidence of an increased risk among individuals under age 25, with the greatest risk among 6–17-year-olds (intention-to-treat RR = 2.90, 95% CI = 1.72, 4.91; per-protocol RR = 3.34, 95% CI = 1.59, 7.00). Our finding of an increased suicidal behaviour risk among individuals under age 25 reflects evidence from RCTs. We found no evidence of an effect in the high-risk group of individuals with past suicidal behaviour. Further studies with information on a wider array of confounders are called for.

ObjectiveExamine the effect of tamoxifen and aromatase inhibitors (AIs) on the risk of 12 clinically relevant cardiovascular outcomes in postmenopausal female breast cancer survivors.MethodsWe carried out two prospective cohort studies among postmenopausal women with breast cancer in UK primary care and hospital data (2002–2016) and US Surveillance, Epidemiology and End Results-Medicare data (2008–2013). Using Cox adjusted proportional hazards models, we compared cardiovascular risks between AI and tamoxifen users; and in the USA, between users of both drug classes and women receiving no endocrine therapy.Results10 005 (UK) and 22 027 (USA) women with postmenopausal breast cancer were included. In both countries, there were higher coronary artery disease risks in AI compared with tamoxifen users (UK age-standardised incidence rate: 10.17 vs 7.51 per 1000 person-years, HR: 1.29, 95% CI 0.94 to 1.76; US age-standardised incidence rate: 36.82 vs 26.02 per 1000 person-years, HR: 1.29, 95% C I1.06 to 1.55). However, comparisons with those receiving no endocrine therapy (US data) showed no higher risk for either drug class and a lower risk in tamoxifen users (age-standardised incidence rate tamoxifen vs unexposed: 26.02 vs 35.19 per 1000 person-years, HR: 0.74, 95% 0.60 to 0.92; age-standardised incidence rate AI vs unexposed: 36.82 vs 35.19, HR: 0.96, 95% CI 0.83 to 1.10). Similar patterns were seen for other cardiovascular outcomes (arrhythmia, heart failure and valvular heart disease). As expected, there was more venous thromboembolism in tamoxifen compared with both AI users and those unexposed.ConclusionsHigher risks of several cardiovascular outcomes among AI compared with tamoxifen users appeared to be driven by protective effects of tamoxifen, rather than cardiotoxic effects of AIs.

While subgroup analyses are common in epidemiologic research, restriction to subgroup members can yield imprecise estimates. We aimed to demonstrate how methods extending inferences to external targets improve precision of subgroup estimates under the major assumption effects differ between subgroup members and nonmembers due to measured effect measure modifiers (EMMs) and membership is independent of the effect after conditioning on EMMs. We applied this approach in the Panitumumab Randomized Trial in Combination with Chemotherapy for Metastatic Colorectal Cancer to Determine Efficacy. Assuming Hispanic vs non-Hispanic ethnicity was independent of the effect conditional on measured EMMs, we weighted non-Hispanic White participants to resemble Hispanic participants in EMMs, assigned Hispanic participants weights of 1, and estimated weighted 9-month progression-free survival differences (PFSDs) with 95% confidence limits from 2000 bootstraps. We also explored outcome-based approaches. Finally, we examined a situation where the method generates biased estimates (targeting participants with mutant-type Kirsten rat sarcoma virus (KRAS), which determines efficacy). While the Hispanic participant-only analysis estimated a 9-month panitumumab PFSD of −7.1% (95% CI −32%, 19%), the weighted combined estimate targeting Hispanic participants was much more precise (−3.7%, 95% CI: −16%, 9.2%). Other analytic approaches yielded similar results. Meanwhile, the weighted combined estimate targeting mutant-type KRAS participants appeared biased (−2.2%, 95% CI: −7.5%, 3.3%) vs the subgroup-only estimate (−11%, 95% CI: −18%, −2.3%). While extending inferences from study populations to specific targets can improve the precision of estimates in small subgroups, violating key assumptions creates bias for many subgroups of interest. Understanding the benefits and harms in specific subgroups of patients is an important part of epidemiologic and public health research. Unfortunately, commonly used methods to do subgroup analyses can result in estimates with lots of uncertainty. Repurposing methods that have traditionally been used to “generalize” or “transport” effect estimates from specific studies to the types of patients more likely to be encountered in the real world could be used to obtain more informative estimates in subgroups without ignoring differences between different types of patients. In this project, we applied this strategy to the Panitumumab Randomized Trial in Combination with Chemotherapy for Metastatic Colorectal Cancer to Determine Efficacy (PRIME) to create much less variable estimates of the treatment effect in Hispanic participants without ignoring the fact that there were more Hispanic participants with a tumor variation that changed the effect of treatment. On the other hand, when we tried to apply this strategy to improve estimates in patients with that tumor variation, we ended up with a misleading effect estimate. While these methods can reduce uncertainty about the benefits of treatment in specific subgroups interesting to researchers, they can result in incorrect subgroup estimates when their assumptions are violated. [Display omitted] •Methods extending inferences to targets can lower subgroup analyses’ variance.•These methods should only be used when key assumptions about heterogeneity are met.•We demonstrate benefits and risks of the approach using simulated and real data.•Research estimating effects in some small subgroups may benefit from these methods.•Estimates resulting from these methods should be interpreted with caution.

Investigators, patients, or clinicians knowing which treatment is assigned in pragmatic randomised trials and observational analyses can lead to detection bias (ie, systematic differences in determining outcomes between groups). A structural definition of detection bias with directed acyclic graphs is provided, together with several published examples. Why negative control outcomes are best placed to assess detection bias is discussed, and how to correctly select a negative control outcome for this purpose is explained.

BackgroundWe aimed at estimating the causal effect of switching from precarious to standard employment on the 6-year and 12-year risk of all-cause mortality among workers aged 20-55 years in Sweden.MethodsWe emulated a series of 12 target trials starting every year between 2005 and 2016 using Swedish register data (n=251 273). We classified precariously employed individuals using a multidimensional approach at baseline as (1) remaining in precarious employment (PE) (73.8%) and (2) shifting to standard employment (26.2%). All-cause mortality was measured from 2006 to 2017. We pooled data for all 12 emulated trials and used covariate-adjusted pooled logistic regression to estimate intention-to-treat and per-protocol effects via risk ratios (RRs) and standardised risk curves (the parametric g-formula).ResultsShifting from precarious to standard employment decreases the 12-year risk of death by 20% on the relative scale (RR: 0.82, 95% CI: 0.73; 0.93), regardless of what happens after the initial shift. However, we estimated a 12-year risk reduction of 30% on the relative scale for workers shifting from precarious to standard employment and staying within this employment category for the full 12 years (RR: 0.71, 95% CI: 0.54; 0.95).ConclusionsThis study finds that shifting from low to higher-quality employment conditions (ie, stable employment, sufficient income levels and high coverage by collective agreements) decreases the risk of death. Remaining in PE increases the risk of premature mortality. Our results emphasise the necessity of ensuring decent work for the entire working population to accomplish the 2030 Agenda for Sustainable Development.