Using Negative Control Populations to Assess Unmeasured Confounding and Direct Effects.

Sometimes treatment effects are absent in a subgroup of the population. For example, penicillin has no effect on severe symptoms in individuals infected by resistant Staphylococcus aureus , and codeine has no effect on pain in individuals with certain polymorphisms in the CYP2D6 enzyme. Subgroups where a treatment is ineffective are often called negative control populations or placebo groups. They are leveraged to detect bias in different disciplines. Here we present formal criteria that justify the use of negative control populations to rule out unmeasured confounding and mechanistic (direct) causal effects. We further argue that negative control populations, satisfying our formal conditions, are available in many settings, spanning from clinical studies of infectious diseases to epidemiologic studies of public health interventions. Negative control populations can also be used to rule out placebo effects in unblinded randomized experiments. As a case study, we evaluate the effect of mobile stroke unit dispatches on functional outcomes at discharge in individuals with suspected stroke, using data from a large trial. Our analysis supports the hypothesis that mobile stroke units improve functional outcomes in these individuals.

Re-evaluating the mythical divide between traditional and novel cardiovascular risk factors in rheumatoid arthritis.

Cardiovascular (CV) risk factors for rheumatoid arthritis (RA) are conventionally classified as 'traditional' and 'novel'. We argue that this classification is obsolete and potentially counterproductive. Further, we discuss problems with the common practice of adjusting for traditional CV risk factors in statistical analyses. These analyses do not target well-defined effects of RA on CV risk. Ultimately, we propose a future direction for cardiorheumatology research that prioritises optimising current treatments and identifying novel therapeutic targets over further categorisation of well-known risk factors.

Estimation of separable direct and indirect effects in continuous time.

Many research questions involve time-to-event outcomes that can be prevented from occurring due to competing events. In these settings, we must be careful about the causal interpretation of classical statistical estimands. In particular, estimands on the hazard scale, such as ratios of cause-specific or subdistribution hazards, are fundamentally hard to interpret causally. Estimands on the risk scale, such as contrasts of cumulative incidence functions, do have a clear causal interpretation, but they only capture the total effect of the treatment on the event of interest; that is, effects both through and outside of the competing event. To disentangle causal treatment effects on the event of interest and competing events, the separable direct and indirect effects were recently introduced. Here we provide new results on the estimation of direct and indirect separable effects in continuous time. In particular, we derive the nonparametric influence function in continuous time and use it to construct an estimator that has certain robustness properties. We also propose a simple estimator based on semiparametric models for the two cause-specific hazard functions. We describe the asymptotic properties of these estimators and present results from simulation studies, suggesting that the estimators behave satisfactorily in finite samples. Finally, we reanalyze the prostate cancer trial from Stensrud et al. (2020).

Causal inference in continuous time: an example on prostate cancer therapy.

In marginal structural models (MSMs), time is traditionally treated as a discrete parameter. In survival analysis on the other hand, we study processes that develop in continuous time. Therefore, Røysland (2011. A martingale approach to continuous-time marginal structural models. Bernoulli 17, 895-915) developed the continuous-time MSMs, along with continuous-time weights. The continuous-time weights are conceptually similar to the inverse probability weights that are used in discrete time MSMs. Here, we demonstrate that continuous-time MSMs may be used in practice. First, we briefly describe the causal model assumptions using counting process notation, and we suggest how causal effect estimates can be derived by calculating continuous-time weights. Then, we describe how additive hazard models can be used to find such effect estimates. Finally, we apply this strategy to compare medium to long-term differences between the two prostate cancer treatments radical prostatectomy and radiation therapy, using data from the Norwegian Cancer Registry. In contrast to the results of a naive analysis, we find that the marginal cumulative incidence of treatment failure is similar between the strategies, accounting for the competing risk of other death.

Sufficient Cause Interaction for Time-to-event Outcomes.

Methods to assess sufficient cause interactions are well developed for binary outcomes. We extend these methods to handle time-to-event outcomes, which occur frequently in medicine and epidemiology. Based on theory for marginal structural models in continuous time, we show how to assess sufficient cause interaction nonparametrically, allowing for censoring and competing risks. We apply the method to study interaction between intensive blood pressure therapy and statin treatment on all-cause mortality.

The surprising implications of familial association in disease risk.

BACKGROUND: A wide range of diseases show some degree of clustering in families; family history is therefore an important aspect for clinicians when making risk predictions. Familial aggregation is often quantified in terms of a familial relative risk (FRR), and although at first glance this measure may seem simple and intuitive as an average risk prediction, its implications are not straightforward. METHODS: We use two statistical models for the distribution of disease risk in a population: a dichotomous risk model that gives an intuitive understanding of the implication of a given FRR, and a continuous risk model that facilitates a more detailed computation of the inequalities in disease risk. Published estimates of FRRs are used to produce Lorenz curves and Gini indices that quantifies the inequalities in risk for a range of diseases. RESULTS: We demonstrate that even a moderate familial association in disease risk implies a very large difference in risk between individuals in the population. We give examples of diseases for which this is likely to be true, and we further demonstrate the relationship between the point estimates of FRRs and the distribution of risk in the population. CONCLUSIONS: The variation in risk for several severe diseases may be larger than the variation in income in many countries. The implications of familial risk estimates should be recognized by epidemiologists and clinicians.

Inequality in genetic cancer risk suggests bad genes rather than bad luck.

Heritability is often estimated by decomposing the variance of a trait into genetic and other factors. Interpreting such variance decompositions, however, is not straightforward. In particular, there is an ongoing debate on the importance of genetic factors in cancer development, even though heritability estimates exist. Here we show that heritability estimates contain information on the distribution of absolute risk due to genetic differences. The approach relies on the assumptions underlying the conventional heritability of liability model. We also suggest a model unrelated to heritability estimates. By applying these strategies, we describe the distribution of absolute genetic risk for 15 common cancers. We highlight the considerable inequality in genetic risk of cancer using different metrics, e.g., the Gini Index and quantile ratios which are frequently used in economics. For all these cancers, the estimated inequality in genetic risk is larger than the inequality in income in the USA.

Diastolic hypotension due to intensive blood pressure therapy: Is it harmful?

BACKGROUND AND AIMS: Reducing the diastolic blood pressure (DBP) below a certain threshold may lead to inadequate organ perfusion. This raises some concerns, because pharmacotherapy reduces both systolic and diastolic pressure. We aimed to investigate whether a pathway from intensive systolic blood pressure (SBP) treatment influences cardiovascular outcomes by inducing too low DBP. METHODS: We had access to data from the Systolic Blood Pressure Intervention Trial (SPRINT) including 9361 patients with a SBP of 130 mmHg or higher and an increased cardiovascular risk. In a formal mediation analysis, we investigated whether the effect of intense (target SBP: 120 mm Hg) vs. standard (target SBP: 140 mmHg) intervention on a composite endpoint would be mediated through an indirect, potentially harmful, effect through low DBP (< 60 mmHg). RESULTS: Adjusting for treatment, we find that low DBP per se is associated with poor cardiovascular outcomes (HR 1.90 (95% CI [1.46, 2.47]). However, in a formal mediation analyses, we observed that the unadjusted indirect effect of intensive blood pressure treatment going through low DBP of HR 1.12 (95% CI [1.06, 1.18]) attenuates to a statistically non-significant effect of HR 1.04 (95% CI [0.98, 1.10]) after adjustment for important covariates, suggesting that the mere association is considerably confounded. CONCLUSIONS: The increased risk in subjects with diastolic pressure below 60 cannot fully be explained by the intensive treatment itself, but may be due to other measured factors. More generally, this analysis shows that adjusting for mediator-outcome confounding is essential, even in RCTs.