Clinical Research Services
Methodological approaches organised by study design. Each study-type page provides a complete analytical framework, strategies for addressing common reviewer questions, and a deliverables list.
Illustrative study types · 10 methodological figures · No actual client projects represented
01 Clinical Research
Browse analytical frameworks, methodological risks and deliverables by study design.
Illustrative data. All figures and numbers on this page are Illustrative data (synthetic), generated by Yanto Health using R simulation scripts. They do not represent any actual patients, hospitals or client projects. Generation scripts are available on request.
Methodological approaches and common reviewer queries for four common study designs. All figures use Illustrative data to show how results can be presented.
Propensity score matching (PSM) + multivariable Cox model, with IPTW for sensitivity analysis.
Figure 1 · Illustrative data · Synthetic illustrative data Illustrative Kaplan-Meier curves and Love plot using the same Illustrative data as /case-cohort/.
View methodology details
PRISMA 2020 flow + heterogeneity I² and Egger's test + indirect comparisons through network meta-analysis.
Figure 2 · Illustrative data · Synthetic illustrative data Illustrative forest and funnel plots using the same Illustrative data as /case-meta/.
View methodology details
ROC discrimination + calibration curves + decision curve analysis (DCA, assessing clinical utility) + bootstrap internal validation.
Figure 3 · Illustrative data · Synthetic illustrative data Illustrative ROC and calibration curves using the same Illustrative data as /case-diagnostic/.
View methodology details
Multiple imputation + time-stratified Cox + time-varying covariates to address immortal time bias.
Figure 4 · Illustrative data · Synthetic illustrative data Illustrative landmark-stratified hazard ratios and sensitivity analysis summary using the same Illustrative data as /case-rws/.
View methodology detailsIllustrations are generated by the same R script (seed 20260519), with values matching the corresponding example pages.
Illustrative data from a multicentre retrospective cohort · N=2,840 · 60-month follow-up window. Core methods: propensity score matching (PSM) + Cox proportional hazards models + inverse probability of treatment weighting (IPTW) sensitivity analysis. Focus: preventing immortal time bias, directed acyclic graph (DAG)-guided confounding adjustment, competing risks and E-value reporting.
Illustrative data · K=18 randomised controlled trials (RCTs) · I²=43% · Includes indirect comparisons using network meta-analysis. Core methods: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 + four-database search + Risk of Bias (RoB) 2.0 assessment + meta-regression for heterogeneity + Egger / Trim-fill publication bias assessment + Grading of Recommendations Assessment, Development and Evaluation (GRADE) evidence quality assessment.
Illustrative data · Training set 845 + external validation set 360 · External validation area under the curve (AUC)=0.87 · Least absolute shrinkage and selection operator (LASSO) + 1000 bootstrap resamples. Core methods: intraclass correlation coefficient (ICC) feature stability screening + strict training/validation separation to prevent data leakage + LASSO regularisation for dimensionality reduction + decision curve analysis (DCA) + item-by-item Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis, including Artificial Intelligence (TRIPOD+AI) checks.
Illustrative data · 8,200 initially at risk · 4 health insurance databases · Cox model with time-varying covariates. Core methods: standardised International Classification of Diseases (ICD) code mapping + precise timeline alignment to prevent immortal time bias + multiple imputation by chained equations (MICE) + time-varying Cox + negative control outcome sensitivity analysis.
02 Clinical Research
✓ indicates commonly used methods listed in the source; suitability depends on the specific research question.
| Analytical method | Cohort | Meta META | Diagnostic model | Real-world data (RWD) |
|---|---|---|---|---|
| DAG causal diagram | ✓ Core | — | — | ✓ Core |
| PSM / IPTW propensity scores | ✓ Primary analysis | — | — | ✓ Sensitivity |
| Cox / time-varying Cox | ✓ Primary model | — | — | ✓ Time-varying format |
| LASSO / bootstrap validation | — Optional | — | ✓ Core | — |
| MICE multiple imputation | ✓ Routine | — | ✓ Within training set | ✓ Extensive missingness |
| GRADE / RoB assessment | — | ✓ Core | — TRIPOD+AI | — |
| DCA decision curves | — | — | ✓ Clinical value | — |
| Competing risks models | ✓ As appropriate | — | — | ✓ Routine |
03 Clinical Research
The following are illustrative methodological formulations organised by reporting guideline items. They do not reproduce actual reviewer comments or represent any particular journal or reviewer.
Immortal time bias from exposure starting after follow-up; failure to report standardised mean differences (SMDs) after PSM; unaddressed competing events; and untruncated IPTW weights. See the relevant example page for full preventive strategies.
High I² without meta-regression to explore its sources; searching only 2 databases; single-reviewer screening without Kappa reporting; unclear GRADE downgrading; and funnel plot asymmetry without small-study effect assessment and sensitivity analyses.
Data leakage from feature selection before data splitting (inflated training AUC); no independent external validation set; reporting AUC without DCA; and no ICC assessment of feature stability.
Mixed ICD-9/10 versions without a mapping strategy; unaddressed immortal time bias; using only baseline values for time-varying covariates; no assessment of between-cohort consistency before pooling databases; and no negative control outcome analysis.
04 Clinical Research
For projects with completed data collection requiring statistical analysis, covering PSM / Cox / meta-analysis / LASSO / MICE.
Support before data collection: refining Population, Intervention, Comparator and Outcome (PICO) + sample size estimation + prespecified statistical analysis plan (SAP), controlling methodological risks before database lock.
Compare Collaboration starter (from ¥12,800), Full-service management (from ¥25,800) and Multicentre / RCT (custom pricing).
Get started · Methodology assessment
Submit your research question, available data and intended outputs. Reply within 24 hours (business days first), with feasibility feedback, viable statistical approaches and a recommended package in our first response.
Our involvement in a paper is disclosed truthfully: where our staff meet the authorship criteria of the International Committee of Medical Journal Editors (ICMJE), they are listed as authors under our company's real affiliation; any other contribution is stated in the acknowledgements. Acceptance is decided by the journal editorial office and peer review. We do not promise publication outcomes.
