CoChat

A Guide to Handling Missing Data in Cost-Effectiveness Analysis Conducted Within Randomised Controlled Trials

Open access
Faria, Rita, Gomes, Manuel, Epstein, David, White, Ian R. · 2014 · PharmacoEconomics

The DOI matches its Crossref record; 46 of the 49 references it cites were found in a registry or an index; none of the 28 sources its claims cite is checked for support yet.

Abstract

Missing data are a frequent problem in cost-effectiveness analysis (CEA) within a randomised controlled trial. Inappropriate methods to handle missing data can lead to misleading results and ultimately can affect the decision of whether an intervention is good value for money. This article provides practical guidance on how to handle missing data in within-trial CEAs following a principled approach: (i) the analysis should be based on a plausible assumption for the missing data mechanism, i.e. whether the probability that data are missing is independent of or dependent on the observed and/or unobserved values; (ii) the method chosen for the base-case should fit with the assumed mechanism; and (iii) sensitivity analysis should be conducted to explore to what extent the results change with the assumption made. This approach is implemented in three stages, which are described in detail: (1) descriptive analysis to inform the assumption on the missing data mechanism; (2) how to choose between alternative methods given their underlying assumptions; and (3) methods for sensitivity analysis. The case study illustrates how to apply this approach in practice, including software code. The article concludes with recommendations for practice and suggestions for future research.

Overview

CoChat Summary

## TL;DR

This paper gives a practical, three-stage guide for handling missing data in within-trial cost-effectiveness analysis: diagnose the likely missingness mechanism, choose a method consistent with that assumption, and then test sensitivity to departures from MAR. It argues that ad hoc approaches such as complete-case analysis, mean imputation, and last-value-carried-forward are usually inadequate for costs and QALYs.

## Key findings

  • The recommended workflow is: (1) descriptive analysis to inform the missing-data mechanism, (2) choose a base-case method that matches that mechanism, and (3) run sensitivity analyses for alternative assumptions.
  • In the REFLUX case study, complete-data proportions fell over follow-up but were similar between arms: 75% to 65% in the surgery group and 82% to 63% in the medical-management group from year 1 to year 5.
  • In that case study, the main analyses gave different point estimates depending on method: incremental costs were £1,668 (CCA), £1,305 (MI), and £1,338 (mixed model); incremental QALYs were 0.301, 0.244, and 0.227, respectively.
  • The sensitivity analysis in the REFLUX study found the surgery cost-effectiveness result was robust to most plausible MNAR departures; the probability of cost-effectiveness changed materially only when QALYs for missing-data individuals in the surgery arm were reduced, and surgery still remained the most likely cost-effective option even if year 1 imputed QALYs were reduced by 50%.

## Methods

  • Methodological guidance paper with a worked illustration using individual patient data from the REFLUX randomized controlled trial, which followed patients for 5 years.
  • The paper combines descriptive missing-data analysis, discussion of alternative handling methods (including MI, IPW, likelihood-based methods, CCA, and available-case analysis), and sensitivity analysis under MNAR assumptions.
  • In the case study, the base case used multiple imputation by chained equations (MI-MICE) under MAR, with a mixed model as an alternative and a pattern-mixture-style sensitivity analysis.

## Limitations

  • The paper states that the true missing-data mechanism is usually unknown, so MNAR cannot be ruled out from the observed data alone.
  • It states that selection models and pattern-mixture approaches can be difficult for non-specialists to implement, and that more research is needed on practical sensitivity-analysis methods and software.
  • Evident from the text: the REFLUX illustration is a single case study, so the numerical results shown there are specific to that trial and are not evidence of general performance across all within-trial CEAs.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1007/s40273-014-0193-3

Cites

46 of 46 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Pattern of missing data. Black shading represents missing data for one or more individuals (arrayed along the horizontal axis) on a particular variable (arrayed along the vertical axis); grey shading represents observed data. a Pattern of missing data on costs. b Pattern of missing data on health-related quality of life (EQ-5D). GP general practitioner
Fig. 1 Pattern of missing data. Black shading represents missing data for one or more individuals (arrayed along the horizontal axis) on a particular variable (arrayed along the vertical axis); grey shading represents observed data. a Pattern of missing data on costs. b Pattern of missing data on health-related quality of life (EQ-5D). GP general practitionerCC BY-NC · PubMed Central
Fig. 2 Comparison of the distribution of imputed values (imputation number 1 to 10) with the observed data (imputation number 0) for quality-adjusted life-years and costs in years 1 and 5. Individual values are represented by dots; the width of a row of dots represents the frequency of values in the distribution. QALYs quality-adjusted life-years
Fig. 2 Comparison of the distribution of imputed values (imputation number 1 to 10) with the observed data (imputation number 0) for quality-adjusted life-years and costs in years 1 and 5. Individual values are represented by dots; the width of a row of dots represents the frequency of values in the distribution. QALYs quality-adjusted life-yearsCC BY-NC · PubMed Central
Fig. 3 Sensitivity analysis: data are missing not at random for QALYs or for costs. Note—imputed costs between year 2 and 5 are increased by 10 %; imputed QALYs between year 2 and 5 are reduced by 10 %. The probability that surgery is cost effective is stable at values close to 1 even if the imputed costs are increased only for the individuals with missing data randomised to the surgery group. Changes in imputed QALYs have an impact on the probability of cost effectiveness if the shift is implemented only in patients with missing data randomised to the surgery group but probability remains above 50 % throughout all scenarios. QALY quality-adjusted life-year
Fig. 3 Sensitivity analysis: data are missing not at random for QALYs or for costs. Note—imputed costs between year 2 and 5 are increased by 10 %; imputed QALYs between year 2 and 5 are reduced by 10 %. The probability that surgery is cost effective is stable at values close to 1 even if the imputed costs are increased only for the individuals with missing data randomised to the surgery group. Changes in imputed QALYs have an impact on the probability of cost effectiveness if the shift is implemented only in patients with missing data randomised to the surgery group but probability remains above 50 % throughout all scenarios. QALY quality-adjusted life-yearCC BY-NC · PubMed Central

Verification

Claims the paper makes

0 Valid · 2 Not valid · 22 Not checked · 81 Own results and interpretations

51 more claims

Create a free account to see all 105 claims and to verify the 28 references still open.

See all claims — free account
  • It may state two separate claims as one. 29
  • Probably not a substantial claim. 24

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 46 references

Cited by

See all 673

Metadata

Authors
Faria, R., Gomes, M., Epstein, D., & White, I. R.
Affiliations
  • University of York (GB)
  • London School of Hygiene & Tropical Medicine (GB)
  • Universidad de Granada (ES)
  • MRC Biostatistics Unit (GB)
Corresponding author
Faria, Rita
Venue
PharmacoEconomics
Published
2014-07-29
Volume
32
Issue
12
Pages
1157-1170
ISSN
  • 1170-7690 (ISSN-L)
  • 1179-2027
OpenAlex
W2151182129
Cited by
673
References
49
Citations by year
2026: 46 · 2025: 92 · 2024: 87 · 2023: 60 · 2022: 82 · 2021: 84 · 2020: 77 · 2019: 44 · 2018: 50 · 2017: 32 · 2016: 12 · 2015: 4 · 2014: 1
FWCI
40.41
Type
journal article
Language
en
Open access
Locations
  • PharmacoEconomics (publishedVersion, cc-by-nc)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Institutional Repository of the University of Granada (University of Granada) (publishedVersion, cc-by-nc)
  • LSHTM Research Online (London School of Hygiene and Tropical Medicine) (acceptedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • LA Referencia (Red Federada de Repositorios Institucionales de Publicaciones Científicas) (submittedVersion, cc-by-nc)
  • UCL Discovery (University College London) (submittedVersion)
Funders
  • University of Aberdeen
  • Medical Research Council (MR/K02177X/1, U1052 60558)
Topics
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Health Systems, Economic Evaluations, Quality of Life
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
Keywords
missing data, cost-effectiveness analysis, randomized controlled trials, missing data mechanisms, sensitivity analysis, value for money, economic evaluation
MeSH terms
Cost-Benefit Analysis, Humans, Statistics as Topic, Randomized Controlled Trials as Topic

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A Guide to Handling Missing Data in Cost-Effectiveness Analysis Conducted Within Randomised Controlled Trials. https://app.cochat.ai/papers/10.1007/s40273-014-0193-3
The paper
Faria, R., Gomes, M., Epstein, D., & White, I. R. (2014). A Guide to Handling Missing Data in Cost-Effectiveness Analysis Conducted Within Randomised Controlled Trials. PharmacoEconomics. https://doi.org/10.1007/s40273-014-0193-3

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.