Concept Architecture
Concept
Theoretically, Individual Patient Data Meta-Analysis (IPD Meta-Analysis) is a meta-analytic approach that synthesises the original participant-level data from multiple studies rather than using published aggregate results. It enables consistent outcome definitions, standardised analyses, adjustment for baseline covariates and investigation of patient-level treatment-effect modifiers. The method exists to provide more accurate and flexible evidence synthesis than conventional aggregate data meta-analysis.
Mathematically, individual patient data meta-analysis applies statistical models directly to pooled participant-level observations using either one-stage or two-stage analytical frameworks. One-stage analyses estimate treatment effects within a single hierarchical regression model, whereas two-stage analyses first estimate study-specific effects before combining them using conventional meta-analysis methods. Random-effects models are frequently employed to account for between-study heterogeneity.
In practice, individual patient data meta-analysis requires collaboration with original investigators to obtain raw study datasets, harmonise variables and perform standardised quality assurance before statistical analysis. It is widely used in health technology assessment, comparative effectiveness research and clinical guideline development when detailed subgroup analyses, time-to-event modelling or adjustment for patient characteristics are required.
Purpose
Used to synthesise participant-level evidence across multiple studies, improve estimation of treatment effects, investigate patient-level effect modifiers and provide robust evidence for clinical and health economic decision-making.
Mathematical Formulae
Primary Formula
One-stage mixed-effects model:
Y?? = ?? + ??X?? + u? + �??
where:
- Y?? = outcome for patient i in study j
- X?? = treatment indicator or covariate
- ?? = overall intercept
- ?? = pooled treatment effect
- u? = random study effect
- �?? = individual-level residual error
Supporting Formulae
Two-stage pooled estimate:
?? = (?w????) / ?w?
Random-effects weight:
w? = 1 / (Var(???) + ��)
Related Mathematical Methods
- Mixed-Effects Regression
- Hierarchical Linear Models
- Random-Effects Meta-Analysis
- Fixed Effect Meta-Analysis
- Cox Proportional Hazards Model
- Logistic Regression
- Meta-Regression
Example
Ten randomised controlled trials evaluating a new anticoagulant contribute individual patient datasets comprising 8,450 participants. A one-stage mixed-effects model estimates an overall hazard ratio of 0.81 (95% confidence interval 0.73 to 0.90) while demonstrating that treatment benefit is greater among patients aged under 75 years. These estimates are subsequently incorporated into a cost-effectiveness model.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| FILTER | =FILTER(A2:H10000,H2:H10000="Included") | Select eligible participant records |
| SORT | =SORT(A2:H10000,1,TRUE) | Organise pooled patient-level data |
| AVERAGEIFS | =AVERAGEIFS(C:C,B:B,"Treatment") | Summarise participant characteristics |
| COUNTIFS | =COUNTIFS(B:B,"Treatment",D:D,"Event") | Calculate event frequencies |
| PivotTable | Participant-level summary | Explore pooled baseline characteristics prior to modelling |
VBA (Optional)
Automate harmonisation, validation and preparation of multiple individual patient datasets before export to statistical software for one-stage or two-stage meta-analysis.
Sources
- Stewart LA, Tierney JF. To IPD or Not to IPD? Advantages and Disadvantages of Systematic Reviews Using Individual Patient Data. Evaluation & the Health Professions. 2002.
- Riley RD, Lambert PC, Abo-Zaid G. Meta-Analysis of Individual Participant Data: Rationale, Conduct and Reporting. BMJ. 2010.
- Higgins JPT, Thomas J, Chandler J, et al. Cochrane Handbook for Systematic Reviews of Interventions.
- NICE. Health Technology Evaluation Manual.
- ISPOR Good Practice Reports.
Related Concepts (2)
Library
Publications
3
Economic Evaluation in Clinical Trials — Glick, Doshi, Sonnad & Polsky, 2nd Edition ed., 2015 (Oxford University Press)
Practical guidance on conducting cost-effectiveness analyses alongside controlled trials, covering trial design, measurement of costs and quality-adjusted life years, handling censored and missing data, and reporting stochastic uncertainty. Volume 4 in the Handbooks in Health Economic Evaluation series.
BookView source →NICE DSU Technical Support Document 6: Embedding Evidence Synthesis in Probabilistic Cost-Effectiveness Analysis — Software Choices — Dias, Welton, Sutton & Ades, TSD 6 ed., 2011 (NICE Decision Support Unit (University of Sheffield))
Guidance on the software options and practical steps for embedding a Bayesian evidence synthesis directly within a probabilistic cost-effectiveness model so that parameter uncertainty is propagated consistently.
NICE DSU Technical Support Document 7: Evidence Synthesis of Treatment Efficacy in Decision Making — A Reviewer’s Checklist — Ades, Caldwell, Reken, Welton, Sutton & Dias, TSD 7 ed., 2011 (NICE Decision Support Unit (University of Sheffield))
A reviewer’s checklist for appraising evidence syntheses of treatment efficacy used in decision making, covering the assumptions and reporting expected of pairwise and network meta-analyses submitted to NICE.
Frequently Asked Questions (6)
What is individual patient data meta-analysis?
A meta-analysis obtaining and analysing original patient-level data from each study, rather than relying only on published summary results.
Source: Stewart & Parmar 1993
What does individual patient data meta-analysis obtain from each study?
Individual patient data meta-analysis obtains the original record for every participant in each included study, rather than the summary figures those studies published. Having the raw patient-level data lets analysts apply consistent definitions across studies, check the original results, and examine how an effect varies with patient characteristics such as age or severity. This depth is what sets it apart, though gathering data from many separate teams is slow and demanding. Working from the raw records is its defining feature. Stewart and Tierney (2002) describe this.
Source: Stewart & Tierney 2002
How does individual patient data meta-analysis work?
Individual patient data meta-analysis works by obtaining the original patient-level datasets from the included studies, checking and standardising them, and analysing them together, either by analysing each study and combining the results or by modelling all the data with appropriate methods that account for the studies. Having the raw data allows consistent outcome definitions, subgroup analyses, and adjustment for patient characteristics. So individual patient data meta-analysis works by collecting and harmonising the raw data across studies and analysing them jointly, using the patient-level information to conduct analyses that are more detailed and consistent than those possible with summary data alone.
Source: Stewart & Parmar 1993
What are the advantages of individual patient data meta-analysis?
The advantages of individual patient data meta-analysis include the ability to standardise outcome definitions and analyses across studies, improving consistency; to examine how effects vary within subgroups and with patient-level characteristics, which summary data cannot support well; to adjust for patient factors; to check the data and update follow-up; and to reduce some reporting biases. These make it more powerful and flexible than aggregate data meta-analysis. So individual patient data meta-analysis is advantageous for the depth, consistency, and flexibility of analysis it allows, particularly for exploring how treatment effects differ among patients, which is why it is often regarded as a gold standard for synthesis.
Source: Stewart & Parmar 1993
What are the challenges of individual patient data meta-analysis?
The challenges of individual patient data meta-analysis include the difficulty and time required to obtain the original datasets from study investigators, who may be unable or unwilling to share them, so that some data may be unavailable, risking bias if the obtained studies are unrepresentative; the effort of checking, harmonising, and analysing the data; and the resources and expertise needed. Data-sharing and confidentiality issues also arise. These challenges mean individual patient data meta-analysis is more demanding and costly than aggregate data meta-analysis, and its feasibility depends on obtaining the raw data, though its analytic advantages often justify the effort where the data can be secured.
Source: Stewart & Parmar 1993
How does individual patient data meta-analysis differ from aggregate data meta-analysis?
Individual patient data meta-analysis obtains and analyses the original patient-level data from each study, while aggregate data meta-analysis combines the summary results reported by each study. The individual patient data approach allows standardised, consistent analysis, examination of subgroups and patient-level effects, and adjustment for patient factors, whereas aggregate data meta-analysis is limited to what is reported and cannot examine patient-level effects directly. However, individual patient data meta-analysis is more demanding, requiring the raw data. So the two differ in the level of data used, with individual patient data offering greater depth and flexibility at the cost of greater effort and feasibility constraints.
Source: Stewart & Parmar 1993
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 3 Dec 2025
Content version: 1.0.0
Canonical Identity
- Term code
- HE-ES-ESM-029
Stable URI · Machine-readable · Resolvable · CC BY 4.0