# causal inference

**URL:** https://discourse.datamethods.org/c/stat/causal/26.md

[Latest](https://discourse.datamethods.org/latest.md) · [Categories](https://discourse.datamethods.org/categories.md) · [Tags](https://discourse.datamethods.org/tags.md)

---

## [About the causal inference category](https://discourse.datamethods.org/t/about-the-causal-inference-category/49)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 0

</div>

Methods and approaches to causal inference

---

## [Causal Formalism and RCTs](https://discourse.datamethods.org/t/causal-formalism-and-rcts/28854)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 27\
**Last updated:** [October 12, 2026, 8:41pm UTC](https://discourse.datamethods.org/t/causal-formalism-and-rcts/28854 "2026-10-12T20:41:48Z")

</div>

Iván Díaz and I have engaged in a long discussion on X.com about the role of causal formalism in randomized clinical trials. This was motivated somewhat by my blog article Causal by Design where I argued that straight-u…

---

## [Feature selection in causal discovery](https://discourse.datamethods.org/t/feature-selection-in-causal-discovery/28735)

<div class="topic-metadata">

**Author:** [@AriRoane](https://discourse.datamethods.org/u/AriRoane)\
**Replies:** 0\
**Last updated:** [May 6, 2026, 3:53pm UTC](https://discourse.datamethods.org/t/feature-selection-in-causal-discovery/28735 "2026-05-06T15:53:56Z")

</div>

Hi! I am using a DAG approach in trying to better specify a research question examining stone-free rate after kidney stone treatment. I have constructed an a priori model using review of relevant literature, but wonderin…

---

## [A High-Fiber Plant-Based Diet in Myeloma Precursor Disorders – Results from the 2 NUTRIVENTION Clinical Trial and Preclinical Vk\*MYC Model](https://discourse.datamethods.org/t/a-high-fiber-plant-based-diet-in-myeloma-precursor-disorders-results-from-the-2-nutrivention-clinical-trial-and-preclinical-vk-myc-model/28559)

<div class="topic-metadata">

**Author:** [@kishtagaria](https://discourse.datamethods.org/u/kishtagaria)\
**Replies:** 2\
**Last updated:** [December 11, 2025, 7:56am UTC](https://discourse.datamethods.org/t/a-high-fiber-plant-based-diet-in-myeloma-precursor-disorders-results-from-the-2-nutrivention-clinical-trial-and-preclinical-vk-myc-model/28559 "2025-12-11T07:56:07Z")

</div>

Shah et al. recently published in Cancer Discovery results from the NUTRIVENTION trial, a single-arm dietary intervention study in patients with myeloma precursor disorders. While the research question is clinically imp…

---

## [Examples of solid causal inferences from purely observational data](https://discourse.datamethods.org/t/examples-of-solid-causal-inferences-from-purely-observational-data/1686)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 27\
**Last updated:** [November 9, 2025, 2:34pm UTC](https://discourse.datamethods.org/t/examples-of-solid-causal-inferences-from-purely-observational-data/1686 "2025-11-09T14:34:53Z")

</div>

I would like to catalog here a few great teaching examples where modern principles of causal inference are used to make solid causality statements from purely observational data. Contributions with brief background, re…

---

## [Causal inferences from RCTs- could “toy” clinical examples promote understanding?](https://discourse.datamethods.org/t/causal-inferences-from-rcts-could-toy-clinical-examples-promote-understanding/18375)

<div class="topic-metadata">

**Author:** [@ESMD](https://discourse.datamethods.org/u/ESMD)\
**Replies:** 17\
**Last updated:** [August 4, 2025, 11:26am UTC](https://discourse.datamethods.org/t/causal-inferences-from-rcts-could-toy-clinical-examples-promote-understanding/18375 "2025-08-04T11:26:35Z")

</div>

Misunderstandings about the types of causal inferences we can make from RCTs abound. They make their way into journals, propagating like viruses and lodging in the brains of other researchers and clinicians. Many have tr…

---

## [Formal Frameworks for Causal Inference -- Preprint by Professor P Richard Hahn](https://discourse.datamethods.org/t/formal-frameworks-for-causal-inference-preprint-by-professor-p-richard-hahn/28107)

<div class="topic-metadata">

**Author:** [@R\_cubed](https://discourse.datamethods.org/u/R_cubed)\
**Replies:** 0\
**Last updated:** [May 16, 2025, 3:42pm UTC](https://discourse.datamethods.org/t/formal-frameworks-for-causal-inference-preprint-by-professor-p-richard-hahn/28107 "2025-05-16T15:42:29Z")

</div>

I follow a number of statisticians on Linkedin; one of them is Professor P. Richard Hahn at Arizona State. He posted a preprint on causal inference and regression adjustment that I’m sure will interest participants in t…

---

## [Lest we forget: “Epidemiology faces its limits”- 30 years later…](https://discourse.datamethods.org/t/lest-we-forget-epidemiology-faces-its-limits-30-years-later/26268)

<div class="topic-metadata">

**Author:** [@ESMD](https://discourse.datamethods.org/u/ESMD)\
**Replies:** 19\
**Last updated:** [April 3, 2025, 1:09pm UTC](https://discourse.datamethods.org/t/lest-we-forget-epidemiology-faces-its-limits-30-years-later/26268 "2025-04-03T13:09:48Z")

</div>

In anticipation of a tsunami of ideologically- and financially (via litigation) - motivated B.S. publications coming out of the U.S., it seems important to unearth this old chestnut from 1995. In the hands of people with…

---

## [Is this patient a "responder"? Essential clinical considerations](https://discourse.datamethods.org/t/is-this-patient-a-responder-essential-clinical-considerations/26657)

<div class="topic-metadata">

**Author:** [@ESMD](https://discourse.datamethods.org/u/ESMD)\
**Replies:** 12\
**Last updated:** [March 11, 2025, 8:01pm UTC](https://discourse.datamethods.org/t/is-this-patient-a-responder-essential-clinical-considerations/26657 "2025-03-11T20:01:27Z")

</div>

Other datamethods threads discuss the pitfalls of “responder analysis” from a statistical standpoint. The purpose of this thread is to highlight the type of clinical evidence that’s needed to label a patient a “responde…

---

## [Does excluding early incident cases help eliminate reverse causality?](https://discourse.datamethods.org/t/does-excluding-early-incident-cases-help-eliminate-reverse-causality/23268)

<div class="topic-metadata">

**Author:** [@JiaqiLi](https://discourse.datamethods.org/u/JiaqiLi)\
**Replies:** 7\
**Last updated:** [November 28, 2024, 2:00am UTC](https://discourse.datamethods.org/t/does-excluding-early-incident-cases-help-eliminate-reverse-causality/23268 "2024-11-28T02:00:18Z")

</div>

In epidemiological cohort studies, it is common to exclude individuals who experience early onset of the disease. For example, in a cohort study examining the association between migraines and the risk of stroke over th…

---

## [ITT analysis under equally high non-adherence in both arms in RCT](https://discourse.datamethods.org/t/itt-analysis-under-equally-high-non-adherence-in-both-arms-in-rct/15396)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 4\
**Last updated:** [June 9, 2024, 10:16am UTC](https://discourse.datamethods.org/t/itt-analysis-under-equally-high-non-adherence-in-both-arms-in-rct/15396 "2024-06-09T10:16:09Z")

</div>

Devereaux et al. ran an RCT comparing dabigatran vs. matched placebo. Patients were supposed to take the drug daily for 2 years. Analyses were done according to the intention-to-treat principle (ITT). Notably, the study…

---

## [Immortal time bias](https://discourse.datamethods.org/t/immortal-time-bias/7149)

<div class="topic-metadata">

**Author:** [@trumanfrancis](https://discourse.datamethods.org/u/trumanfrancis)\
**Replies:** 2\
**Last updated:** [October 29, 2023, 12:27pm UTC](https://discourse.datamethods.org/t/immortal-time-bias/7149 "2023-10-29T12:27:52Z")

</div>

In a retrospective cohort study using an electronic database the exposure is the administration of drug A, the outcome will be mortality. If we select subjects at time zero but subjects that did not receive treatment A …

---

## [Best Causal Inference method for repeated Cross-section Design](https://discourse.datamethods.org/t/best-causal-inference-method-for-repeated-cross-section-design/7080)

<div class="topic-metadata">

**Author:** [@PAVAN\_Pandey](https://discourse.datamethods.org/u/PAVAN_Pandey)\
**Replies:** 1\
**Last updated:** [October 5, 2023, 6:57am UTC](https://discourse.datamethods.org/t/best-causal-inference-method-for-repeated-cross-section-design/7080 "2023-10-05T06:57:58Z")

</div>

Please allow me to introduce myself to you. I Dr Pavan, from India, currently working for an iNGO and as an advisor to the Government of Madhya Pradesh. I am approaching you with a technical problem and seeking your exp…

---

## [RMS Causal Inference](https://discourse.datamethods.org/t/rms-causal-inference/4848)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 51\
**Last updated:** [October 3, 2023, 1:42pm UTC](https://discourse.datamethods.org/t/rms-causal-inference/4848 "2023-10-03T13:42:25Z")

</div>

Regression Modeling Strategies: Causal Inference and Directed Acyclic Graphics This is for questions and discussion about causal inference related to Regression Modeling Strategies. The purposes of these topics are to i…

---

## [Observational data with diagnosis partially determining treatment (imperfect instrumental variable?)](https://discourse.datamethods.org/t/observational-data-with-diagnosis-partially-determining-treatment-imperfect-instrumental-variable/6983)

<div class="topic-metadata">

**Author:** [@martinmodrak](https://discourse.datamethods.org/u/martinmodrak)\
**Replies:** 0\
**Last updated:** [August 8, 2023, 10:39am UTC](https://discourse.datamethods.org/t/observational-data-with-diagnosis-partially-determining-treatment-imperfect-instrumental-variable/6983 "2023-08-08T10:39:53Z")

</div>

Here’s the setting - we have a retrospective dataset of patients undergoing a certain surgery. The surgery has two variants, let’s call them A and B. We are interested in comparing a set of outcomes like length of hospit…

---

## [The relation between Statistical Inference and Causal Inference](https://discourse.datamethods.org/t/the-relation-between-statistical-inference-and-causal-inference/4574)

<div class="topic-metadata">

**Author:** [@R\_cubed](https://discourse.datamethods.org/u/R_cubed)\
**Replies:** 16\
**Last updated:** [June 4, 2023, 10:22pm UTC](https://discourse.datamethods.org/t/the-relation-between-statistical-inference-and-causal-inference/4574 "2023-06-04T22:22:10Z")

</div>

A deep philosophical question came up in the OR vs RR thread. @AndersHuitfeldt wrote: Blockquote My personal view is that trading off bias for precision is “cheating” and gives you uninterpretable inferential statist…

---

## [Longitudinal exposures, causal inference, and proper analyses](https://discourse.datamethods.org/t/longitudinal-exposures-causal-inference-and-proper-analyses/6379)

<div class="topic-metadata">

**Author:** [@lorenzoFabbri](https://discourse.datamethods.org/u/lorenzoFabbri)\
**Replies:** 0\
**Last updated:** [January 23, 2023, 4:54pm UTC](https://discourse.datamethods.org/t/longitudinal-exposures-causal-inference-and-proper-analyses/6379 "2023-01-23T16:54:05Z")

</div>

Setting: Exposures: multiple exposures (biomarkers, all continuous) measured at two time points, once in pregnancy and once in adolescence Outcome: clinical outcomes measured in adolescence (concurrently with measure…

---

## [How to assess an isolated effect or causal relationship in longitudinal data?](https://discourse.datamethods.org/t/how-to-assess-an-isolated-effect-or-causal-relationship-in-longitudinal-data/6173)

<div class="topic-metadata">

**Author:** [@aleksi](https://discourse.datamethods.org/u/aleksi)\
**Replies:** 3\
**Last updated:** [November 7, 2022, 3:23pm UTC](https://discourse.datamethods.org/t/how-to-assess-an-isolated-effect-or-causal-relationship-in-longitudinal-data/6173 "2022-11-07T15:23:39Z")

</div>

My question relates to causal inference in longitudinal data setting. I have a dataset which includes certain measurements from same individuals at two different time points, say at baseline and at 12 months. My main re…

---

## [Selection bias or good data management](https://discourse.datamethods.org/t/selection-bias-or-good-data-management/5881)

<div class="topic-metadata">

**Author:** [@Sudhi\_Upadhyaya](https://discourse.datamethods.org/u/Sudhi_Upadhyaya)\
**Replies:** 6\
**Last updated:** [August 17, 2022, 11:11pm UTC](https://discourse.datamethods.org/t/selection-bias-or-good-data-management/5881 "2022-08-17T23:11:50Z")

</div>

I working with a dataset that is an amalgamation of patients belonging to different cohorts. I have data from 800 patients belonging to Cohort 12345, 350 patient data from Cohort 54321, 900 patient data from Cohort 5689 …

---

## [Confounder related](https://discourse.datamethods.org/t/confounder-related/5823)

<div class="topic-metadata">

**Author:** [@Sudhi\_Upadhyaya](https://discourse.datamethods.org/u/Sudhi_Upadhyaya)\
**Replies:** 2\
**Last updated:** [August 9, 2022, 4:07am UTC](https://discourse.datamethods.org/t/confounder-related/5823 "2022-08-09T04:07:02Z")

</div>

I am working with an Epidemiologist and she said “Why deprivation index, it is only a confounder” ? I am trying to understand what this means and here’s the context behind this statement. The study objective is to esti…

---

## [Causal graph based on data generation process](https://discourse.datamethods.org/t/causal-graph-based-on-data-generation-process/5338)

<div class="topic-metadata">

**Author:** [@al\_stau](https://discourse.datamethods.org/u/al_stau)\
**Replies:** 3\
**Last updated:** [February 22, 2022, 1:05pm UTC](https://discourse.datamethods.org/t/causal-graph-based-on-data-generation-process/5338 "2022-02-22T13:05:21Z")

</div>

Hi, I am quite new to causal inference and want to try some methods for treatment effect estimation. For this purpose, I created a the following data generation process in Python: import numpy as np n = 10000 X3 = np.…

---

## [Cross sectional study design and analysis technique for evaluating impact of COVID 19 on radiotherapy quality of care indicators](https://discourse.datamethods.org/t/cross-sectional-study-design-and-analysis-technique-for-evaluating-impact-of-covid-19-on-radiotherapy-quality-of-care-indicators/4202)

<div class="topic-metadata">

**Author:** [@S\_Chakraborty](https://discourse.datamethods.org/u/S_Chakraborty)\
**Replies:** 1\
**Last updated:** [March 13, 2021, 2:27pm UTC](https://discourse.datamethods.org/t/cross-sectional-study-design-and-analysis-technique-for-evaluating-impact-of-covid-19-on-radiotherapy-quality-of-care-indicators/4202 "2021-03-13T14:27:15Z")

</div>

We are planning a study to investigate if the COVID 19 epidemic impacted the radiotherapy quality of care indicators for cervical cancer patients treated with curative intent radiotherapy. This is a complex treatment whe…

---

## [Reporting with multiple testing](https://discourse.datamethods.org/t/reporting-with-multiple-testing/4157)

<div class="topic-metadata">

**Author:** [@DieterMenne](https://discourse.datamethods.org/u/DieterMenne)\
**Replies:** 6\
**Last updated:** [February 24, 2021, 10:21am UTC](https://discourse.datamethods.org/t/reporting-with-multiple-testing/4157 "2021-02-24T10:21:29Z")

</div>

I have seen the dozens of posting on the multiple-testing subject. I also have done quite a lot of Stan-based analysis and had to defend these against p-value greedy authors and reviewers. I am also co-signatory of the …

---

## [Cause selection while modeling a Graphical Causal Model](https://discourse.datamethods.org/t/cause-selection-while-modeling-a-graphical-causal-model/4060)

<div class="topic-metadata">

**Author:** [@Ramesh\_Iyer](https://discourse.datamethods.org/u/Ramesh_Iyer)\
**Replies:** 3\
**Last updated:** [January 18, 2021, 2:16pm UTC](https://discourse.datamethods.org/t/cause-selection-while-modeling-a-graphical-causal-model/4060 "2021-01-18T14:16:51Z")

</div>

Hi There! I have recently started toying around with Causal Models specifically the kinds from Judea Perl school of thought. In practical application however, I find that there are several numerous causes to an outcome.…

---

## [Collider bias in obesity research](https://discourse.datamethods.org/t/collider-bias-in-obesity-research/3936)

<div class="topic-metadata">

**Author:** [@vildelm](https://discourse.datamethods.org/u/vildelm)\
**Replies:** 0\
**Last updated:** [November 27, 2020, 7:03am UTC](https://discourse.datamethods.org/t/collider-bias-in-obesity-research/3936 "2020-11-27T07:03:30Z")

</div>

Collider bias is common in obesity research and proposed a cause of the several obesity paradoxes shown in epi papers (survival advantage in obese vs lean people with chronic conditions such as heart failure and diabetes…

---

## [Are hazard ratios still hazardous in hierarchical models?](https://discourse.datamethods.org/t/are-hazard-ratios-still-hazardous-in-hierarchical-models/3825)

<div class="topic-metadata">

**Author:** [@hanowell](https://discourse.datamethods.org/u/hanowell)\
**Replies:** 3\
**Last updated:** [October 8, 2020, 2:35pm UTC](https://discourse.datamethods.org/t/are-hazard-ratios-still-hazardous-in-hierarchical-models/3825 "2020-10-08T14:35:40Z")

</div>

Miguel Hernán admonishes against the use of hazard ratios for causal inference because selection bias is built into them: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3653612/ Yet if I build a hierarchical model that al…

---

## [Adherence adjustment in randomized trials is possible, here's why](https://discourse.datamethods.org/t/adherence-adjustment-in-randomized-trials-is-possible-heres-why/137)

<div class="topic-metadata">

**Author:** [@EpiEllie](https://discourse.datamethods.org/u/EpiEllie)\
**Replies:** 4\
**Last updated:** [July 9, 2019, 6:51pm UTC](https://discourse.datamethods.org/t/adherence-adjustment-in-randomized-trials-is-possible-heres-why/137 "2019-07-09T18:51:48Z")

</div>

This post began life as a tweetorial, or twitter tutorial, for International Clinical Trials Day 2018. To read the original tweetorial, click here. The goal of this tweetorial was to explain the evidence for and against …

---

## [Causal inference vs "the usual statistics"](https://discourse.datamethods.org/t/causal-inference-vs-the-usual-statistics/2009)

<div class="topic-metadata">

**Author:** [@jlevy13](https://discourse.datamethods.org/u/jlevy13)\
**Replies:** 1\
**Last updated:** [August 20, 2019, 5:18pm UTC](https://discourse.datamethods.org/t/causal-inference-vs-the-usual-statistics/2009 "2019-08-20T17:18:13Z")

</div>

Do you think that causal inference methodology should be used in all circumstances?

---

## [Does correlation among confounders change amount of bias](https://discourse.datamethods.org/t/does-correlation-among-confounders-change-amount-of-bias/1774)

<div class="topic-metadata">

**Author:** [@Felix\_Thoemmes](https://discourse.datamethods.org/u/Felix_Thoemmes)\
**Replies:** 0\
**Last updated:** [June 21, 2019, 5:28pm UTC](https://discourse.datamethods.org/t/does-correlation-among-confounders-change-amount-of-bias/1774 "2019-06-21T17:28:42Z")

</div>

Quite a while ago I attended a conference talk on sensitivity analysis with multiple confounders. The talk is archived and can be accessed here: http://www.sfu.ca/~lmccandl/LectureACIC.pdf There was one small aspect of…