Bias Assessment of Observational Studies · What can go wrong in an observational study?...
Transcript of Bias Assessment of Observational Studies · What can go wrong in an observational study?...
![Page 1: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/1.jpg)
Bias Assessment of Observational StudiesTHOMAS G. STEWART, PHD
![Page 2: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/2.jpg)
Note: These slides will be posted on my faculty page within days of the conference (after correcting errors and incorporating suggestions made by you).
See tgstewart.xyz
![Page 3: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/3.jpg)
Who am I?• Assistant Professor of Biostatistics at Vanderbilt University Medical Center
• Collaborative experience with a number of national and international registries, most notably• HCV-TARGET• AHSQC
• Collaborative experience with state and national claims and administrative data
![Page 4: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/4.jpg)
Why observational studies?• Ethical considerations:• Can’t randomize individuals to smoke/not-smoke
•Economic considerations:• The data is there … Why not use it?
In the past
More recently
![Page 5: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/5.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
![Page 6: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/6.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
![Page 7: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/7.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
![Page 8: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/8.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
![Page 9: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/9.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Health system preference, insurance policies, physician preference, family history, geography, disease severity, patient choice, calendar time, …
![Page 10: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/10.jpg)
Why care about confounders?If treatment A was only given to patients with severe disease and treatment B was only given to patients with mild disease,
What would one learn from a comparison of outcomes between treatment groups?
Take it to the extreme
![Page 11: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/11.jpg)
Why care about confounders?If treatment A was mostly given to patients with severe disease and treatment B was mostly given to patients with mild disease,
What would one learn from a comparison of outcomes between treatment groups?
Less extreme
![Page 12: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/12.jpg)
Why care about confounders?One of the reasons randomization is so important is that it creates, on average, comparison groups that are balanced in terms of disease severity and other confounders.
What would one learn from a comparison of outcomes between treatment groups created by randomization?
Balance
![Page 13: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/13.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
![Page 14: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/14.jpg)
Disclaimer about examples:I am a statistician, not a clinician.
While I can speak abstractly about the concept of “confounding” and “treatment by indication bias”, I am hesitant to editorialize about specific clinical examples outside the areas of my collaborative experience.
As such, the examples in this workshop of studies that may be biased are drawn from published criticism.
![Page 15: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/15.jpg)
Example:
Criticism:
![Page 16: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/16.jpg)
Criticism:
![Page 17: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/17.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
![Page 18: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/18.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl)
![Page 19: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/19.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl)
Eligible for both arms
![Page 20: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/20.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
![Page 21: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/21.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Just because a patient received treatment A, was the patient eligible to receive treatment B?
![Page 22: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/22.jpg)
Why care about eligibility?If patients in treatment group A were only eligible to receive treatment A, and patients in treatment group B was only eligible to receive treatment,
What would one learn from a comparison of outcomes between treatment groups?
Take it to the extreme
![Page 23: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/23.jpg)
Why care about eligibility?One of the reasons an interventional comparison is meaningful is that all patients could have potentially have received both treatments. There was some degree of equipoise of the treatments for the study subjects.
What would one learn from a comparison of outcomes between treatment groups created by randomization?
Equipoise
![Page 24: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/24.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
![Page 25: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/25.jpg)
Example from collaborative experience:Open repair vs Laparoscopic repair of ventral hernias
Among surgeons in a national registry of hernia repairs, certain complex hernias are only repaired using an open surgical approach.
To include patients with this specific type of hernia in a comparison of open and laparoscopic approaches would be a mistake because this group of patients is not eligible to receive both surgical approaches.
![Page 26: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/26.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
At risk for outcomes
![Page 27: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/27.jpg)
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none
???
Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
At risk for outcomes?Prospectively at risk
![Page 28: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/28.jpg)
Why care about at-risk?Are males reasonable controls for studies of treatment for ovarian cancer?Take it to the extreme
![Page 29: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/29.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Prospectively at risk Inappropriate controls
![Page 30: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/30.jpg)
Editorial:Selection of appropriate controls is more difficult than most than most of us (including me) realize.
Selection of appropriate controls is especially difficult when the control arm is passive, i.e., defined by NOT receiving an intervention.
![Page 31: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/31.jpg)
Editorial:Selection of appropriate controls is more difficult than most than most of us (including me) realize.
Selection of appropriate controls is especially difficult when the control arm is passive, i.e., defined by NOT receiving an intervention.
Why? The decision to not actively intervene is generally not recorded in the medical record. It is also correlated with patient priorities and the desire for aggressive care.
![Page 32: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/32.jpg)
Editorial:Selection of appropriate controls is more difficult than most than most of us (including me) realize.
Selection of appropriate controls is especially difficult when the control arm is passive, i.e., defined by NOT receiving an intervention.
Why? The decision to not actively intervene is generally not recorded in the medical record. It is also correlated with patient priorities and the desire for aggressive care.
Are these passive controls really eligible to receive both treatment arms? What is the “entry date” for passive controls?
![Page 33: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/33.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Prospectively at risk Inappropriate controls
![Page 34: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/34.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Prospectively at risk Inappropriate controls
Identical follow-up protocol
![Page 35: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/35.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Prospectively at risk Inappropriate controls
Identical follow-up protocol ???, EHR, Registry,
![Page 36: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/36.jpg)
Why care about follow-up?If patients in treatment group A interact with their health-care providers weekly, but patients in treatment group B interact with their care providers only as needed,
Are treatment groups equally likely to capture study endpoints?
Take it to the extreme
![Page 37: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/37.jpg)
Failure to consider equipoise
What can go wrong in an observational study?
Observational
Population
Endpoint
Exposure
Confounders
Data capture
RCT
Randomization/Intervention
On average, none Failure to address confounding
Prospective (Incl/Excl) Retrospectively got exposure
Prospectively at risk Inappropriate controls
Identical follow-up protocol Inconsistent follow-up
![Page 38: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/38.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
![Page 39: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/39.jpg)
Design Choices vs Statistical MethodsNot every bias can be corrected with a statistical method. Some bias is only controlled with appropriate study design.
![Page 40: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/40.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
Regression, propensity score methods
![Page 41: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/41.jpg)
Matching (The beautiful table 1 approach)
![Page 42: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/42.jpg)
![Page 43: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/43.jpg)
Mean differences in observed covariates are minimized in the matched cohort.
![Page 44: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/44.jpg)
Matching selects pairs (or groups) of subjects that are similar.
Some subjects are discarded from the analysis.
![Page 45: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/45.jpg)
A common approach to identify similar subjects between each treatment group is to use the propensity score.
The propensity score is a summary measure of all important covariates. It is an estimate of the probability of receiving one of the treatments.
![Page 46: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/46.jpg)
Regression (Use all the data approach)
𝐸 OUTCOME −] = 𝛽, + 𝛽. 𝑇𝑅𝐸𝐴𝑇𝑀𝐸𝑁𝑇 𝐴 + 𝛽4 𝑀𝐴𝐿𝐸 + 𝛽6 𝐴𝐺𝐸 + 𝛽8 𝐴𝐺𝐸4 + …
![Page 47: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/47.jpg)
Regression (Use all the data approach)
𝐸 OUTCOME −] = 𝛽, + 𝛽. 𝑇𝑅𝐸𝐴𝑇𝑀𝐸𝑁𝑇 𝐴 + 𝛽4 𝑀𝐴𝐿𝐸 + 𝛽6 𝐴𝐺𝐸 + 𝛽8 𝐴𝐺𝐸4 + …
Association of interest Potential confounding variables
![Page 48: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/48.jpg)
Regression (Use all the data approach)
𝐸 OUTCOME −] = 𝛽, + 𝛽. 𝑇𝑅𝐸𝐴𝑇𝑀𝐸𝑁𝑇 𝐴 + 𝛽4 𝑀𝐴𝐿𝐸 + 𝛽6 𝐴𝐺𝐸 + 𝛽8 𝐴𝐺𝐸4 + …
Association of interest Potential confounding variables
METFORMIN
![Page 49: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/49.jpg)
Regression and Propensity Score methods(Belt and suspenders approaches)
𝐸 OUTCOME −] = 𝛽, + 𝛽. TRT A + 𝛽4 MALE + 𝛽6 AGE + 𝛽8 AGE4 + …+ 𝑅𝐶𝑆(𝑃𝑆)
Association of interest
Potential confounding variables Flexible association with propensity score
![Page 50: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/50.jpg)
Regression and Propensity Score methods(Belt and suspenders approaches)
1. Regression with propensity score as flexible covariate
2. Propensity score matching then regression
3. Propensity score weighting with regression
![Page 51: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/51.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
Regression, propensity score methods, OTHER
![Page 52: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/52.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
Covariate overlap plots
Regression, propensity score methods
![Page 53: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/53.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
Covariate overlap plots
Regression, propensity score methods
Reasonable incl/excl
![Page 54: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/54.jpg)
Editorial:Selection of appropriate controls is more difficult than most than most of us (including me) realize.
Selection of appropriate controls is especially difficult when the control arm is passive, i.e., defined by NOT receiving an intervention.
One solution when the control arm is passive is to use time-varying covariate methods instead of awkward definitions for controls.
![Page 55: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/55.jpg)
Failure to consider equipoise
Ways to address possible pit-falls in observational studies
Pit-fall
Population
Endpoint
Exposure
Confounders
Data capture
Failure to address confounding
Retrospectively got exposure
Inappropriate controls
Inconsistent follow-up
Solution
Covariate overlap plots
Regression, propensity score methods
Reasonable incl/excl
Design
![Page 56: Bias Assessment of Observational Studies · What can go wrong in an observational study? Observational Population Endpoint Exposure Confounders Data capture RCT. What can go wrong](https://reader034.fdocuments.us/reader034/viewer/2022052612/5f09e5487e708231d42904b9/html5/thumbnails/56.jpg)
Questions?