Evidence-based Medicine
Review: Target Trial Emulation in Observational Studies
24 Nov, 2024 | 19:43h | UTCIntroduction:
Randomized controlled trials (RCTs) are the gold standard for evaluating medical interventions but are often resource-intensive. Observational studies using real-world data (RWD) from electronic health records (EHRs) or claims data offer advantages like larger sample sizes and timely availability. However, they are prone to biases. “Target trial emulation” aligns observational studies with hypothetical RCTs to improve validity and causal inference.
Key Recommendations:
- Design Alignment with RCTs: Observational studies should replicate key trial design elements like eligibility, treatment strategies, follow-up, and outcome measures. This reduces biases and clarifies treatment-effect definitions.
- Specification of “Time Zero”: Clearly define the start of eligibility, treatment assignment, and follow-up to prevent immortal time and selection bias. Proper alignment ensures comparable treatment groups.
- Addressing Bias and Confounding: Observational studies remain vulnerable to confounding from unmeasured variables. Researchers must use matching, sensitivity analyses, and robust methods to assess findings’ reliability.
- Data Quality Challenges: Using RWD requires careful validation of completeness, consistency, and accuracy. Missing or inconsistent data can affect results. Clear descriptions of data sources and quality metrics are essential.
- Best Practices:
- Protocols and SAPs: Pre-specify protocols and statistical plans (SAPs) to enhance transparency and mitigate bias.
- Clear Reporting: State the observational nature of the study to avoid confusion with RCTs. Use tables to outline target trial elements.
- Sensitivity Analyses: Explore unmeasured confounding and perform negative control outcomes to ensure robustness.
Limitations:
Despite emulating RCTs, observational studies cannot fully eliminate biases inherent in nonrandomized designs. Confounding, selection bias, and data quality issues remain risks. Careful design, pre-specification, and transparency are essential for credible results.
Conclusion:
Target trial emulation enhances observational study design by aligning it with RCT principles, mitigating biases, and improving the reliability of findings. However, it is not a substitute for randomization. Rigorous implementation and interpretation are critical for improving patient care and outcomes.
Reference:
Hubbard RA, Gatsonis CA, Hogan JW, et al. “Target Trial Emulation” for Observational Studies — Potential and Pitfalls. New England Journal of Medicine. 2024 Nov 23. DOI: http://doi.org/10.1056/NEJMp2407586
Systematic Review | Poor handling of continuous predictors in clinical prediction models using logistic regression
11 Aug, 2023 | 15:19h | UTC
Perspective | Creation and adoption of large language models in medicine
9 Aug, 2023 | 15:38h | UTCCreation and Adoption of Large Language Models in Medicine – JAMA (free for a limited period)
Commentary: Rethinking large language models in medicine – Stanford Medicine
Perspective | Artificial-intelligence search engines wrangle academic literature
9 Aug, 2023 | 15:33h | UTCArtificial-intelligence search engines wrangle academic literature – Nature
Research Letter | GPT-3.5 and GPT-4 show low accuracy in citing journal articles
9 Aug, 2023 | 15:31h | UTCAccuracy of Chatbots in Citing Journal Articles – JAMA Network Open
Commentary on Twitter
21% of medical journal articles cited by GPT-4 were found to be fake; GPT-3.5 cited an estimated 98% fake articles. Narrower topics had more fake articles than broader topics. Despite its promise, ChatGPT is currently not a reliable source of medical data. https://t.co/DCTIkT1OkZ
— JAMA Network Open (@JAMANetworkOpen) August 9, 2023
Survey | The manifold costs of being a non-native English speaker in science
7 Aug, 2023 | 15:00h | UTCThe manifold costs of being a non-native English speaker in science – PLOS Biology
Commentary from the author: Non-native English speaking scientists work much harder just to keep up, global research reveals – The Conversation
News Releases:
The cost of being a non-native English speaker in science – PLOS
Science language barrier could cost countless careers – University of Queensland
Commentary on Twitter
Some potential solutions to reducing disadvantages for non-native English speakers in publishing & conferences
By @tatsuya_amano & colleagues from their seminal @PLOSBiology paperhttps://t.co/luBndai3FQ @PLOS pic.twitter.com/68x6rZBMW4
— Madhu Pai, MD, PhD (@paimadhu) August 5, 2023
Review | ‘Pragmatic’ trials changing research: Balancing traditional controls with real-world care needs
4 Aug, 2023 | 12:08h | UTCClinical Trials Overview: From Explanatory to Pragmatic Clinical Trials – Mayo Clinic Proceedings
Perspective | The most important trial in modern cardiology (Critical appraisal of the ISCHEMIA trial)
2 Aug, 2023 | 14:11h | UTCChapter 3: The Most Important Trial in Modern Cardiology – Sensible Medicine
See also:
Chapter 1: What Does it Mean to Discover a Coronary Blockage? – By Dr John Mandrola
Chapter 2: On the meaning of a coronary blockage – By Dr John Mandrola
Related/Original Articles:
ISCHEMIA Trial: Initial Invasive vs. Conservative Strategy for Stable Coronary Disease
ISCHEMIA Trial: Management of Coronary Disease in Patients with Advanced Kidney Disease
Review | Futility in clinical trials
1 Aug, 2023 | 14:32h | UTCFutility in Clinical Trials – JAMA (free for a limited period)
Opinion | Bad science needs to be called out: the medical profession loses trust when bad studies are used to promote policy
1 Aug, 2023 | 14:29h | UTCBad Science Needs to Be Called Out – Sensible Medicine
Related: Did Republicans die more during the pandemic bc they didn’t get the vax? – By Dr. Vinay Prassad
Original Study: Excess Death Rates for Republican and Democratic Registered Voters in Florida and Ohio During the COVID-19 Pandemic – JAMA Internal Medicine
Critical Appraisal | Did Republicans die more during the pandemic bc they didn’t get the vax?
1 Aug, 2023 | 14:28h | UTCDid Republicans die more during the pandemic bc they didn’t get the vax? – By Dr. Vinay Prassad
Related: Bad Science Needs to Be Called Out – Sensible Medicine
Original Study: Excess Death Rates for Republican and Democratic Registered Voters in Florida and Ohio During the COVID-19 Pandemic – JAMA Internal Medicine
BMJ Investigation | Medical royal colleges receive millions from drug and medical devices companies
28 Jul, 2023 | 14:19h | UTCMedical royal colleges receive millions from drug and medical devices companies – The BMJ
News Release: Medical royal colleges receive millions from drug/medical device companies – BMJ Newsroom
Commentary on Twitter
Drug and medical device companies have given £9m to UK royal colleges in recent years.
Scroll through our series of #BMJInfographics to see how much individual companies have paid each college 👇 https://t.co/s03xSl18BP pic.twitter.com/PX1Lzmtb4N
— The BMJ (@bmj_latest) July 27, 2023
JAMA guidance for authors, peer reviewers, and editors on use of AI, language models, and chatbots
28 Jul, 2023 | 14:18h | UTCGuidance for Authors, Peer Reviewers, and Editors on Use of AI, Language Models, and Chatbots – JAMA
See also, just published: New recommendations of the International Committee of Medical Journal Editors: use of artificial intelligence – European Heart Journal
Related
ChatGPT et al? Not So Fast, Say Journal Editors – TCTMD
Chat GPT will change Medicine – Vinay Prasad’s Observations and Thoughts
Can artificial intelligence help for scientific writing? – Critical Care
Artificial intelligence in academic writing: a paradigm-shifting technological advance
Perspective | Generating scholarly content with ChatGPT: ethical challenges for medical publishing
ChatGPT and the future of medical writing (ChatGPT itself wrote this paper)
ChatGPT is fun, but not an author – Science
Tools such as ChatGPT threaten transparent science; here are our ground rules for their use – Nature
ChatGPT listed as author on research papers: many scientists disapprove – Nature
Abstracts written by ChatGPT fool scientists
New Cochrane handbook for systematic reviews of diagnostic test accuracy
25 Jul, 2023 | 14:02h | UTCCochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy – Cochrane Library
News Release: Introducing the new Cochrane Handbook for Diagnostic Test Accuracy – Cochrane Library
Perspective | Medicine is plagued by untrustworthy clinical trials. How many studies are faked or flawed?
24 Jul, 2023 | 13:12h | UTCMedicine is plagued by untrustworthy clinical trials. How many studies are faked or flawed? – Nature
Related: Dozens of recent clinical trials may contain wrong or falsified data (free study by John Carlisle)
Commentary on Twitter
.@Nature reports on the investigations that suggest that, in some fields, at least one-quarter of clinical trials might be problematic or even entirely made up. The researchers behind these investigations are calling for stronger scrutiny. https://t.co/uztDC6JwbH
— Nature Portfolio (@NaturePortfolio) July 22, 2023
Pooled Analysis | High disease prevalence linked to increased sensitivity, decreased specificity
24 Jul, 2023 | 12:51h | UTC
Correspondence | Potential “healthy vaccinee bias” in a study of BNT162b2 vaccine against Covid-19
21 Jul, 2023 | 13:43h | UTC
Commentary from the author on Twitter (thread – click for more)
Our correspondence is out in @njem written w @VPrasadMDMPH & Ram Duriseti
We show v strong evidence of healthy vaccinee bias in an Israeli study of Pfizer booster vs covid death, which wasn't disclosed
Could explain ALL reported benefit of the booster🧵https://t.co/F3WeafvUnw pic.twitter.com/b29YQ0lblh
— Tracy Høeg, MD, PhD (@TracyBethHoeg) July 20, 2023
A guide and pragmatic considerations for applying GRADE to network meta-analysis
12 Jul, 2023 | 13:44h | UTCA guide and pragmatic considerations for applying GRADE to network meta-analysis – The BMJ
Commentary on Twitter
How can you assess the certainty of evidence from network meta-analyses using the GRADE approach?
A new guide with practical strategies 👇https://t.co/g2S4a17GM7 pic.twitter.com/FHZOSDN0b8
— The BMJ (@bmj_latest) July 2, 2023
Cohort Study | Fewer than half of new drugs add substantial therapeutic value over existing treatments
11 Jul, 2023 | 14:01h | UTCNews Release: Fewer than half of new drugs add substantial therapeutic value over existing treatments – BMJ Newsroom
Editorial: Patients need better treatments, not just more of the same – The BMJ
GRADE guidance 36: updates to GRADE’s approach to addressing inconsistency
5 Jul, 2023 | 01:05h | UTC
Scoping Review | High correlation of positive outcomes in oncology meta-analyses with industry funding
3 Jul, 2023 | 14:21h | UTC
Perspective | Ensuring ethical postprogression therapy for patients in randomized trial control arms
28 Jun, 2023 | 13:16h | UTC
Commentary on Twitter
Ensuring Ethical Postprogression Therapy for Patients in RCT Arms
"Patients participating in clinical trials make significant sacrifices, and in return the academic and clinical communities owe them optimal treatment after progression"@JCO_ASCO https://t.co/2dv2rky2WN pic.twitter.com/rQfRx3T4Ii
— Yakup Ergün (@dr_yakupergun) June 22, 2023
Editorial | Who should judge treatment effects as unimportant?
22 Jun, 2023 | 15:05h | UTCWho should judge treatment effects as unimportant? – Journal of Physiotherapy
Brief Review | Selection bias due to conditioning on a collider
21 Jun, 2023 | 13:32h | UTCSelection bias due to conditioning on a collider – The BMJ
Perspective | Benefits and pitfalls of social media in cardiology
16 Jun, 2023 | 14:13h | UTC
Commentary on Twitter
Medical #education has entered the #DigitalTransformation era 🤖
What is the role of #SoMe and multimedia platforms in #cardiology❓
Here is our view from the @JACCJournals #JACCCaseReports #SoMe Editors ❤️#CardioTwitter #MedEd #CardioEd
➡️Link to full paper below🧵 pic.twitter.com/0W77HzoJ6Q
— Alessandro Beneduce (@ABeneduceMD) June 9, 2023