Assessing Heterogeneity of Treatment Effect in Real-World Data

Segal, Jodi B; Varadhan, Ravi; Groenwold, Rolf H H; Li, Xiaojuan; Nomura, Kaori; Kaplan, Sigal; Ardeshir-Rouhani-Fard, Shirin; Heyward, James; Nyberg, Fredrik; Burcu, Mehmet

doi:10.7326/m22-1510

Cited by 14 publications

(9 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“… 46 The potential heterogeneity of treatment effects will be explored via the causal forest method. 47 …”

Section: Discussionmentioning

confidence: 99%

Protocol for the Chinese Real-World Evidence for Acute Spinal Cord Injury (ChiRES) study: a prospective, observational, multicentre cohort study of acute spinal cord injury

Yuan,

Sun,

et al. 2024

BMJ Open

View full text Add to dashboard Cite

IntroductionSpinal cord injury (SCI) is a catastrophic event with devastating physical, social and occupational consequences for patients and their families. The number of patients with acute SCI in China continues to grow rapidly, but there have been no large prospective cohort studies of patients with acute SCI. This proposed study aims to establish a multicentre, extensive sample cohort of clinical data and biological samples of patients in China, which would aid the systematisation and standardisation of clinical research and treatment of acute SCI, thus reducing the heavy burden of acute SCI on patients and society.Methods and analysisThe Chinese Real-World Evidence for Acute Spinal Cord Injury (ChiRES) study is an observational, multicentre cohort study of patients with acute SCI admitted to the Qilu Hospital of Shandong University and other participating centres with prospective collection of their clinical data and biological samples. We aim to recruit 2097 patients in this study. Demographics, disease history, emergency intervention information, motor and sensory examinations, surgical information, medication information and rehabilitation evaluation will be recorded. This will facilitate the development of a prediction model for complications and prognosis of patients with acute SCI and an evaluation of the current management of acute SCI. Among these variables, detailed information on surgical treatment will also be used to assess procedures for acute SCI treatment. Outcome measurements, including the International Standard for Neurological Classification of Spinal Cord Injury examinations, the occurrence of complications and death, will be performed repeatedly during follow-up. We will analyse imaging data and blood samples to develop SCI imaging markers and biomarkers.Ethics and disseminationThis study protocol has been approved by the Medical Ethics Committee of the Qilu Hospital of Shandong University and all other participating centres. The findings will be disseminated in peer-reviewed journals and academic conferences.

show abstract

“… 46 The potential heterogeneity of treatment effects will be explored via the causal forest method. 47 …”

Section: Discussionmentioning

confidence: 99%

Protocol for the Chinese Real-World Evidence for Acute Spinal Cord Injury (ChiRES) study: a prospective, observational, multicentre cohort study of acute spinal cord injury

Yuan,

Sun,

et al. 2024

BMJ Open

View full text Add to dashboard Cite

show abstract

“…Heterogeneity of treatment effect occurs when the response to an intervention differs systematically across individuals with different characteristics. 22 , 23 , 24 , 25 Intrinsic biological characteristics of a participant (eg, diagnosis, fetal sex, genetics) or extrinsic factors (eg, socioeconomic factors) may systematically alter response to a therapy. 22 , 23 , 24 , 25 Qualitative heterogeneity, where the direction of response varies within subgroups, complicates interpretation of an overall effect because the results for the component subgroups tend to cancel when pooled across participants.…”

Section: Discussionmentioning

confidence: 99%

Progesterone for Neurodevelopment in Fetuses With Congenital Heart Defects

Gaynor,

Moldenhauer,

Zullo

et al. 2024

JAMA Netw Open

View full text Add to dashboard Cite

ImportanceNeurodevelopmental outcomes for children with congenital heart defects (CHD) have improved minimally over the past 20 years.ObjectivesTo assess the feasibility and tolerability of maternal progesterone therapy as well as the magnitude of the effect on neurodevelopment for fetuses with CHD.Design, Setting, and ParticipantsThis double-blinded individually randomized parallel-group clinical trial of vaginal natural progesterone therapy vs placebo in participants carrying fetuses with CHD was conducted between July 2014 and November 2021 at a quaternary care children’s hospital. Participants included maternal-fetal dyads where the fetus had CHD identified before 28 weeks’ gestational age and was likely to need surgery with cardiopulmonary bypass in the neonatal period. Exclusion criteria included a major genetic or extracardiac anomaly other than 22q11 deletion syndrome and known contraindication to progesterone. Statistical analysis was performed June 2022 to April 2024.InterventionParticipants were 1:1 block-randomized to vaginal progesterone or placebo by diagnosis: hypoplastic left heart syndrome (HLHS), transposition of the great arteries (TGA), and other CHD diagnoses. Treatment was administered twice daily between 28 and up to 39 weeks’ gestational age.Main Outcomes and MeasuresThe primary outcome was the motor score of the Bayley Scales of Infant and Toddler Development-III; secondary outcomes included language and cognitive scales. Exploratory prespecified subgroups included cardiac diagnosis, fetal sex, genetic profile, and maternal fetal environment.ResultsThe 102 enrolled fetuses primarily had HLHS (n = 52 [50.9%]) and TGA (n = 38 [37.3%]), were more frequently male (n = 67 [65.7%]), and without genetic anomalies (n = 61 [59.8%]). The mean motor score differed by 2.5 units (90% CI, −1.9 to 6.9 units; P = .34) for progesterone compared with placebo, a value not statistically different from 0. Exploratory subgroup analyses suggested treatment heterogeneity for the motor score for cardiac diagnosis (P for interaction = .03) and fetal sex (P for interaction = .04), but not genetic profile (P for interaction = .16) or maternal-fetal environment (P for interaction = .70).Conclusions and RelevanceIn this randomized clinical trial of maternal progesterone therapy, the overall effect was not statistically different from 0. Subgroup analyses suggest heterogeneity of the response to progesterone among CHD diagnosis and fetal sex.Trial RegistrationClinicalTrials.gov Identifier: NCT02133573

show abstract

“…Regardless of the analysis frameworks used in clinical data mining, researchers cannot randomize patients to receive treatment [ 123 , 124 ]; rather, the patients have tendentiously received the treatments in past clinical practice [ 125 ]. Unfortunately, in most cases, this specific information on treatment selection is also missed or unrecorded in EHRs for clinical data mining.…”

Section: Evaluating the Consequences Of Controlling Confounders On Th...mentioning

confidence: 99%

Clinical data mining: challenges, opportunities, and recommendations for translational applications

Qiao,

Chen,

Qian

et al. 2024

J Transl Med

View full text Add to dashboard Cite

Clinical data mining of predictive models offers significant advantages for re-evaluating and leveraging large amounts of complex clinical real-world data and experimental comparison data for tasks such as risk stratification, diagnosis, classification, and survival prediction. However, its translational application is still limited. One challenge is that the proposed clinical requirements and data mining are not synchronized. Additionally, the exotic predictions of data mining are difficult to apply directly in local medical institutions. Hence, it is necessary to incisively review the translational application of clinical data mining, providing an analytical workflow for developing and validating prediction models to ensure the scientific validity of analytic workflows in response to clinical questions. This review systematically revisits the purpose, process, and principles of clinical data mining and discusses the key causes contributing to the detachment from practice and the misuse of model verification in developing predictive models for research. Based on this, we propose a niche-targeting framework of four principles: Clinical Contextual, Subgroup-Oriented, Confounder- and False Positive-Controlled (CSCF), to provide guidance for clinical data mining prior to the model's development in clinical settings. Eventually, it is hoped that this review can help guide future research and develop personalized predictive models to achieve the goal of discovering subgroups with varied remedial benefits or risks and ensuring that precision medicine can deliver its full potential.

show abstract

Assessing Heterogeneity of Treatment Effect in Real-World Data

Cited by 14 publications

References 57 publications

Protocol for the Chinese Real-World Evidence for Acute Spinal Cord Injury (ChiRES) study: a prospective, observational, multicentre cohort study of acute spinal cord injury

Protocol for the Chinese Real-World Evidence for Acute Spinal Cord Injury (ChiRES) study: a prospective, observational, multicentre cohort study of acute spinal cord injury

Progesterone for Neurodevelopment in Fetuses With Congenital Heart Defects

Clinical data mining: challenges, opportunities, and recommendations for translational applications

Contact Info

Product

Resources

About