Choosing sensitivity analyses for randomised trials: principles
Open accessAbstract
BACKGROUND: Sensitivity analyses are an important tool for understanding the extent to which the results of randomised trials depend upon the assumptions of the analysis. There is currently no guidance governing the choice of sensitivity analyses. DISCUSSION: We provide a principled approach to choosing sensitivity analyses through the consideration of the following questions: 1) Does the proposed sensitivity analysis address the same question as the primary analysis? 2) Is it possible for the proposed sensitivity analysis to return a different result to the primary analysis? 3) If the results do differ, is there any uncertainty as to which will be believed? Answering all of these questions in the affirmative will help researchers to identify relevant sensitivity analyses. Treating analyses as sensitivity analyses when one or more of the answers are negative can be misleading and confuse the interpretation of studies. The value of these questions is illustrated with several examples. SUMMARY: By removing unreasonable analyses that might have been performed, these questions will lead to relevant sensitivity analyses, which help to assess the robustness of trial results.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
When the findings from a sensitivity analysis are consistent with the primary results, confidence that the primary results are valid can be increased.
“When the findings from a sensitivity analysis are consistent with the primary results, confidence that the primary results are valid can be increased.445 447”
From SPIRIT 2025 explanation and elaboration: updated guideline for protocols of randomised trials (2025) · cites this paper as [445]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Schafer, J. L. (1999) · Statistical Methods in Medical Research · doi:10.1177/096228029900800102 · Cited by 3,949 · cited 1× · see citing works
“Single imputation fails to allow for the uncertainty induced by missing data, and does not in general lead to valid inferences [12].”
BackgroundRiociguat for the treatment of pulmonary arterial hypertension
Ghofrani, H.-A., Galiè, N., Grimminger, F., Grünig, E., Humbert, M., Jing, Z.-C., Keogh, A. M., Langleben, D., Kilama, M. O., Fritsch, A., et al. (2013) · New England Journal of Medicine · doi:10.1056/nejmoa1209655 · Cited by 1,346 · cited 1× · see citing works
“It is common for authors to compare the results of intention-to-treat with per-protocol analysis; see for example [8,9].”
BackgroundScott, R. A. P. (2002) · The Lancet · doi:10.1016/s0140-6736(02)11522-4 · Cited by 1,254 · see citing works
Schafer, J. L. (1999) · Statistical Methods in Medical Research · doi:10.1191/096228099671525676 · Cited by 920 · see citing works
A tutorial on sensitivity analyses in clinical trials: the what, why, when and how
Thabane, L., Mbuagbaw, L., Zhang, S., Samaan, Z., Marcucci, M., Ye, C., Thabane, M., Giangregorio, L., Dennis, B., Kosa, D., et al. (2013) · BMC Medical Research Methodology · doi:10.1186/1471-2288-13-92 · Cited by 893 · cited 5× · see citing works
“Consider the proposal for handling missing data in Table two of Thabane et al [5].”
BackgroundMethodInfluentialStrategy for intention to treat analysis in randomised trials with missing outcome data
White, I. R., Horton, N. J., Carpenter, J. R., statistics, R. I. M. A. S., & Pocock, S. (2011) · BMJ · doi:10.1136/bmj.d40 · Cited by 849 · cited 2× · see citing works
“The sensitivity analyses are reasonable because they: 1) address the same question as the primary analysis, under different assumptions; 2) may or may not lead to different conclusions; and 3) involve different assumptions which may be plausible, although some may be less so: there is genuine uncertainty about the mos…”
BackgroundSelf-citationHu, F. B., Goldberg, J., Hedeker, D., Flay, B. R., & Pentz, M. A. (1998) · American Journal of Epidemiology · doi:10.1093/oxfordjournals.aje.a009511 · Cited by 422 · cited 1× · see citing works
“consider approaches to analysing a study with a longitudinal binary outcome [14], comparing random effects approaches with generalised estimating equations.”
MethodBayesian inference for causal effects in randomized experiments with noncompliance
Imbens, G. W., & Rubin, D. B. (1997) · The Annals of Statistics · doi:10.1214/aos/1034276631 · Cited by 382 · cited 1× · see citing works
“Another analysis investigated the complier average causal effect, which considers what the (average) effect of treatment was in patients who would have adhered to protocol however they were randomised [7].”
BackgroundThe performance of the two‐stage analysis of two‐treatment, two‐period crossover trials
Freeman, P. R. (1989) · Statistics in Medicine · doi:10.1002/sim.4780081202 · Cited by 270 · see citing works
How to Select Covariates to Include in the Analysis of a Clinical Trial
RAAB, G. M., Day, S. J., & Sales, J. (2000) · Controlled Clinical Trials · doi:10.1016/s0197-2456(00)00061-1 · Cited by 185 · cited 1× · see citing works
“Second, choosing the method for the primary analysis based on trial data has been shown to lead to unreliable results in several settings [1-4].”
MethodAssessing potential sources of clustering in individually randomised trials
Kahan, B. C., & Morris, T. P. (2013) · BMC Medical Research Methodology · doi:10.1186/1471-2288-13-58 · Cited by 100 · cited 1× · see citing works
“In general an analysis that accounts for clustering is used because clustering arises through the design [13].”
MethodSelf-citationUses and limitations of randomization-based efficacy estimators
White, I. R. (2005) · Statistical Methods in Medical Research · doi:10.1191/0962280205sm406oa · Cited by 98 · cited 1× · see citing works
“While it is hard to pin down the precise question of per-protocol analysis [10], this is clearly different to the question intention-to-treat addresses.”
BackgroundSelf-citationBias in randomised factorial trials
Kahan, B. C. (2013) · Statistics in Medicine · doi:10.1002/sim.5869 · Cited by 56 · see citing works
Self-citationBendtsen, P., McCambridge, J., Bendtsen, M., Karlsson, N. E., & Nilsén, P. (2012) · Journal of Medical Internet Research · doi:10.2196/jmir.2062 · Cited by 55 · cited 1× · see citing works
“It is common for authors to compare the results of intention-to-treat with per-protocol analysis; see for example [8,9].”
Perkins, G. D., Woollard, M., Cooke, M. W., Deakin, C. D., Horton, J., Lall, R., Lamb, S. E., McCabe, C., Quinn, T., Slowther, A., et al. (2010) · Scandinavian Journal of Trauma Resuscitation and Emergency Medicine · doi:10.1186/1757-7241-18-58 · Cited by 52 · cited 1× · see citing works
“Example The Paramedic trial is designed to assess whether survival of cardiac arrest patients can be improved by equipping ambulances with a mechanical chest compression device, compared to manual chest compressions by the crew [15].”
BackgroundZheng, G., Li, M., Lan, X., Yan, X., Lin, Q., Chen, L., Tao, J., Zheng, X., Li, J., Chen, B., et al. (2013) · Trials · doi:10.1186/1745-6215-14-422 · Cited by 38 · cited 5× · see citing works
“Returning to the missing data example above, if Zheng et al. were to use single (as opposed to multiple) imputation as a sensitivity analysis, as suggested in [5], this could well lead to different conclusions, despite our criticism above that the model for imputation is identical to the model for the primary analysis.”
BackgroundMethodInfluentialPauw, B. E. D., Sable, C. A., Walsh, T. J., Lupinacci, R. J., Bourque, M. R., Wise, B. A., Nguyen, B., DiNubile, M. J., & Teppler, H. (2006) · Transplant Infectious Disease · doi:10.1111/j.1399-3062.2006.00127.x · Cited by 32 · see citing works
The consequences of proportional hazards based model selection
Campbell, H., & Dean, C. B. (2013) · Statistics in Medicine · doi:10.1002/sim.6021 · Cited by 22 · cited 1× · see citing works
“Second, choosing the method for the primary analysis based on trial data has been shown to lead to unreliable results in several settings [1-4].”
MethodCarnes, D., Taylor, S. J., Homer, K., Eldridge, S. M., Bremner, S. A., Pincus, T., Rahman, A., & Underwood, M. R. (2013) · BMJ Open · doi:10.1136/bmjopen-2012-002492 · Cited by 16 · cited 1× · see citing works
“The Copers trial [18] is designed to evaluate a selfmanagement course for patients with chronic musculoskeletal pain.”
Background
Flagged first, then by citations
Cited by
Sorted by citations
Jakobsen, J. C., Gluud, C., Wetterslev, J., & Winkel, P. (2017) · BMC Medical Research Methodology · doi:10.1186/s12874-017-0442-1 · Cited by 2,747 · see citing works
“Sensitivity analyses may show how assumptions, different from those made in the primary analysis influence the results obtained [3, 6].”
BackgroundAir Versus Oxygen in ST-Segment–Elevation Myocardial Infarction
Stub, D. A., Smith, K. L., Bernard, S., Nehme, Z., Stephenson, M., Bray, J., Cameron, P. A., Barger, B., Ellims, A., Taylor, A. J., et al. (2015) · Circulation · doi:10.1161/circulationaha.114.014494 · Cited by 534 · see citing works
“Definitions of the end points used in this study are provided in Table I in the online-only Data Supplement.”
MethodCONSORT 2025 explanation and elaboration: updated guideline for reporting randomised trials
Hopewell, S., Chan, A.-W., Collins, G. S., Hróbjartsson, A., Moher, D., Schulz, K. F., Tunn, R., Aggarwal, R., Berkwits, M., Berlin, J. A., et al. (2025) · BMJ · doi:10.1136/bmj-2024-081124 · Cited by 205 · see citing works
Self-citationCro, S., Morris, T. P., Kenward, M. G., & Carpenter, J. R. (2020) · Statistics in Medicine · doi:10.1002/sim.8569 · Cited by 203 · see citing works
Self-citationThe estimands framework: a primer on the ICH E9(R1) addendum
Kahan, B. C., Hindley, J., Edwards, M., Cro, S., & Morris, T. P. (2024) · BMJ · doi:10.1136/bmj-2023-076316 · Cited by 180 · see citing works
Self-citationBhatnagar, R., Piotrowska, H., Laskawiec-Szkonter, M., Kahan, B. C., Luengo-Fernández, R., Pepperell, J., Evison, M., Holme, J., Al-Aloul, M., Psallidas, I., et al. (2019) · JAMA · doi:10.1001/jama.2019.19997 · Cited by 160 · see citing works
“(%) 46/161 (29) 44/159 (28) Odds ratio, 1.”
Self-citationSPIRIT 2025 explanation and elaboration: updated guideline for protocols of randomised trials
Hróbjartsson, A., Boutron, I., Hopewell, S., Moher, D., Schulz, K. F., Collins, G. S., Tunn, R., Aggarwal, R., Berkwits, M., Berlin, J. A., et al. (2025) · BMJ · doi:10.1136/bmj-2024-081660 · Cited by 171 · see citing works
Self-citationRehal, S., Morris, T. P., Fielding, K., Carpenter, J. R., & Phillips, P. P. J. (2016) · BMJ Open · doi:10.1136/bmjopen-2016-012594 · Cited by 154 · see citing works
“80% 6 (36) 19 (31) 85% 11 (7) 5 (8) 90% 65 (39) 26 (43) 71–99% (excluding the above) 21 (12) 11 (18) Not reported/unclear 10 (6) 0 Composite outcome Yes 78 (46) 37 (61) No 90 (54) 24 (39) Disease Heart disease 30 (18) 13 (21) Blood disorder 19 (11) 6 (10) Cancer 16 (10) 8 (13) Diabetes 11 (7) 2 (3)”
Self-citationReve, E. V. H., & de Bruin, E. D. (2014) · BMC Geriatrics · doi:10.1186/1471-2318-14-134 · Cited by 132 · see citing works
“The results of sensitivity analyses [93,97] in addition to the primary intention to treat analyses where outliers are included, and the participants are analysed in the group where they were initially allocated, are reported in an additional file of this manuscript (see Additional file 1).”
BackgroundPlanning a method for covariate adjustment in individually randomised trials: a practical guide
Morris, T. P., Walker, A. S., Williamson, E. J., & White, I. R. (2022) · Trials · doi:10.1186/s13063-022-06097-z · Cited by 97 · see citing works
“This prompts sensitivity analysis [29].”
MethodSelf-citationTaylor, R. W., Gray, A. R., Heath, A. M., Galland, B. C., Lawrence, J., Sayers, R., Healey, D. M., Tannock, G. W., Meredith-Jones, K. A., Hanna, M., et al. (2018) · American Journal of Clinical Nutrition · doi:10.1093/ajcn/nqy090 · Cited by 82 · see citing works
A four-step strategy for handling missing outcome data in randomised trials affected by a pandemic
Cro, S., Morris, T. P., Kahan, B. C., Cornelius, V. R., & Carpenter, J. R. (2020) · BMC Medical Research Methodology · doi:10.1186/s12874-020-01089-6 · Cited by 57 · see citing works
“Sensitivity analysis should address the same question as the primary analysis [32].”
BackgroundSelf-citationSensitivity Analysis: A Method to Promote Certainty and Transparency in Nursing and Health Research
Mowbray, F., Manlongat, D., & Shukla, M. (2022) · Canadian Journal of Nursing Research · doi:10.1177/08445621221107108 · Cited by 53 · see citing works
“Data screening, cleaning, and analysis, commonly reveal unanticipated barriers and findings, further highlighting the value of post-hoc sensitivity analyses (Morris et al., 2014).”
BackgroundInfluentialTaylor, S. J., Carnes, D., Homer, K., Pincus, T., Kahan, B. C., Hounsome, N., Eldridge, S., Spencer, A., Diaz-Ordaz, K., Rahman, A., et al. (2016) · Programme Grants for Applied Research · doi:10.3310/pgfar04140 · Cited by 44 · see citing works
Self-citationAlper, B. S., Foster, G., Thabane, L., Rae-Grant, A. D., Malone-Moses, M., & Manheimer, E. W. (2020) · BMJ Evidence-Based Medicine · doi:10.1136/bmjebm-2020-111386 · Cited by 43 · see citing works
Vorland, C. J., Brown, A. W., Dawson, J., Dickinson, S., Golzarri‐Arroyo, L., Hannon, B. A., Heo, M., Heymsfield, S. B., Jayawardene, W. P., Kahathuduwa, C. N., et al. (2021) · International Journal of Obesity · doi:10.1038/s41366-021-00909-z · Cited by 40 · see citing works
“However, simply comparing per protocol to ITT analyses as a sensitivity analysis is suboptimal;they estimate different things [113].”
Kahan, B. C., Forbes, G., & Cro, S. (2020) · BMC Medicine · doi:10.1186/s12916-020-01706-7 · Cited by 37 · see citing works
Self-citationSensitivity analysis in clinical trials: three criteria for a valid sensitivity analysis
Parpia, S., Morris, T. P., Phillips, M. R., Wykoff, C. C., Steel, D. H., Thabane, L., Bhandari, M., Chaudhary, V., for the Retina Evidence Trials InterNational Alliance (R.E.T.I.N.A.) Study Group, Sivaprasad, S., et al. (2022) · Eye · doi:10.1038/s41433-022-02108-0 · Cited by 32 · see citing works
“proposed a framework to conduct such analysis [2] and suggest that a particular analysis can be classified as a sensitivity analysis if it meets the following criteria: (1) the proposed analysis aims to answer the same question as to the primary analysis, (2) there is a possibility that the proposed analysis will lead…”
ResultSelf-citationLeurent, B., Gomes, M., Cro, S., Wiles, N., & Carpenter, J. R. (2019) · Health Economics · doi:10.1002/hec.3963 · Cited by 32 · see citing works
“…each assumption is likely to be a matter of debate, but it is important to keep in mind that the true missing data mechanism is always unknown and that the aim of the sensitivity analyses is to indicate how results could differ under a range of plausible assumptions (Morris, Kahan, & White, 2014).”
Reporting of randomized factorial trials was frequently inadequate
Kahan, B. C., Tsui, M., Jairath, V., Scott, A. M., Altman, D. G., Beller, E., & Elbourne, D. R. (2019) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2019.09.018 · Cited by 30 · see citing works
Self-citationMissing data: Issues, concepts, methods
Pham, T. M., Pandis, N., & White, I. R. (2024) · Seminars in Orthodontics · doi:10.1053/j.sodo.2024.01.007 · Cited by 25 · see citing works
“25 − 27 What are ad-hoc methods for dealing with missing values and are they valid?”
MethodSelf-citationMcKenzie, J. E., Taljaard, M., Hemming, K., Arnup, S. J., Giraudeau, B., Eldridge, S. M., Hooper, R., Kahan, B. C., Li, T., Moher, D., et al. (2025) · BMJ · doi:10.1136/bmj-2024-080472 · Cited by 30 · see citing works
Self-citation(2024) · doi:10.2196/47515 · Cited by 24 · see citing works
“Under the MAR assumption, consistent findings between the primary analysis and sensitivity analysis can strengthen the reliability of the findings [61-64], at least in situations where both the primary and sensitivity analyses are expected to be valid under similar assumptions (eg, multiple…”
BackgroundPalliative clinical trials in advanced chronic liver disease: Challenges and opportunities
Verma, S., Hingwala, J. P., Low, J., Patel, A., Verma, M., Bremner, S. A., Haddadin, Y., Shinall, M. C., Komenda, P. V. J., & Ufere, N. N. (2023) · Journal of Hepatology · doi:10.1016/j.jhep.2023.06.018 · Cited by 24 · see citing works
Alencar, C. D. M., de Paula, B. L. F., Araújo, J. L. N., Alves, E. B., Jassé, F. F. D. A., & Silva, C. M. (2018) · Journal of Esthetic and Restorative Dentistry · doi:10.1111/jerd.12386 · Cited by 23 · see citing works
Missing data in bioarchaeology II: A test of ordinal and continuous data imputation
Wissler, A., Blevins, K. E., & Buikstra, J. E. (2022) · American Journal of Biological Anthropology · doi:10.1002/ajpa.24614 · Cited by 22 · see citing works
Best practice for analysis of shared clinical trial data
Hollis, S., Fletcher, C., Lynn, F., Urban, H.-J., Branson, J., Burger, H.-U., Smith, C. T., Sydes, M. R., & Gerlinger, C. (2016) · BMC Medical Research Methodology · doi:10.1186/s12874-016-0170-y · Cited by 22 · see citing works
“Methodology has evolved in recent years, for example Morris et al [19] and Mallinckrodt et al [20] give good examples of dealing with missing data assumptions within a sensitivity analyses framework.”
MethodDetailed statistical analysis plan for the Danish Palliative Care Trial (DanPaCT)
Johnsen, A. T., Petersen, M. A., Gluud, C. N., Lindschou, J., Fayers, P. M., Sjøgren, P., Pedersen, L., Neergaard, M. A., Vejlgaard, T. B., Damkier, A., et al. (2014) · Trials · doi:10.1186/1745-6215-15-376 · Cited by 20 · see citing works
“Sensitivity analyses will be made to test the robustness of the conclusion according to the following rules: a) they should address the same questions as the primary analysis; b) it should be a possibility that they will arrive at another conclusion; and c) if another conclusion is reached, there shall be genuine unce…”
BackgroundMultiple imputation and test‐wise deletion for causal discovery with incomplete cohort data
Witte, J., Foraita, R., & Didelez, V. (2022) · Statistics in Medicine · doi:10.1002/sim.9535 · Cited by 20 · see citing works
Staudt, A., Freyer‐Adam, J., Ittermann, T., Meyer, C., Bischof, G., John, U., & Baumann, S. (2022) · BMC Medical Research Methodology · doi:10.1186/s12874-022-01727-1 · Cited by 20 · see citing works
“Rather, sensitivity analyses are warranted to examine if the conclusions drawn from a particular study differ, depending on whether an ignorable (MAR) or non-ignorable (MNAR) missing data mechanism is assumed [20, 33].”
BackgroundPlow, M. A., Motl, R. W., Finlayson, M. L., & Béthoux, F. (2020) · Journal of Telemedicine and Telecare · doi:10.1177/1357633x20964693 · Cited by 19 · see citing works
“Because little consensus exists on what are the optimal methods to examine response heterogeneity, three common methods – data visualization, moderated regression analysis and responder analysis – were selected to be consistent with the principles of sensitivity analyses.(27) Data visualization involved developing bar…”
MethodThe Sense and Sensibility of Sensitivity Analyses
Cheng, D. M., & Hogan, J. W. (2024) · New England Journal of Medicine · doi:10.1056/nejmp2403318 · Cited by 15 · see citing works
Rehal, S., Cro, S., Phillips, P. P. J., Fielding, K., & Carpenter, J. R. (2023) · Clinical Trials · doi:10.1177/17407745231176773 · Cited by 14 · see citing works
“When missing outcome data arises, it is acknowledged that analysis should not only consist of a primary analysis under the most plausible missing data assumption but also include sensitivity analysis, under alternative contextually plausible assumptions concerning the distribution of the unobserved data.(14) In practi…”
Estimands for factorial trials
Kahan, B. C., Morris, T. P., Goulão, B., & Carpenter, J. R. (2022) · Statistics in Medicine · doi:10.1002/sim.9510 · Cited by 11 · see citing works
Self-citationFiorino, G., Sturniolo, G. C., Bossa, F., Cassinotti, A., Sabatino, A. D., Giuffrida, P., & Danese, S. (2019) · Cells · doi:10.3390/cells8060523 · Cited by 12 · see citing works
“Also, sensitivity analysis to assess the effects of missing data was performed [28].”
BackgroundEstimators for handling COVID ‐19‐related intercurrent events with a hypothetical strategy
Lasch, F., & Guizzaro, L. (2022) · Pharmaceutical Statistics · doi:10.1002/pst.2244 · Cited by 11 · see citing works
Xu, J., Wang, Y., He, Q., Xie, S., Feng, S., Xiaofei, W., Wang, W., & Sun, X. (2025) · BMC Medicine · doi:10.1186/s12916-025-04199-4 · Cited by 11 · see citing works
A brief overview of sensitivity analyses
Kunselman, A. R. (2023) · Fertility and Sterility · doi:10.1016/j.fertnstert.2023.04.031 · Cited by 10 · see citing works
INFILTRATION CHARACTERISTICS OF FILM HOLE IRRIGATION UNDER THE INFLUENCE OF MULTIPLE FACTORS
Zhong, Y., Fei, L., & Li, Y. (2020) · Irrigation and Drainage · doi:10.1002/ird.2417 · Cited by 10 · see citing works
“…for different factors, and then the absolute values were obtained to analyse the sensitivity of the soil infiltration rate to the various factors (Morris et al., 2014): ∂i ∂γ ¼ 88:810γ 8:633d0:301T 0:128 (4) ∂i ∂d ¼ 3:502γ 7:633d 0:699T 0:128 (5) Using Equations (4) and (5), the sensitivity of…”
BackgroundMethodKahan, B. C., Doré, C. J., Murphy, M. F., & Jairath, V. (2016) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2016.05.008 · Cited by 10 · see citing works
Self-citationHamel, J. F., Sebille, V., Néel, T. L., Kubis, G., Boyer, F. C., & Hardouin, J.-B. (2015) · Statistical Methods in Medical Research · doi:10.1177/0962280215615158 · Cited by 9 · see citing works
To Aid Scientific Inference, Emphasize Unconditional Descriptions of StatisticsPreprint
Greenland, S., & Rafi, Z. (2019) · arXiv (Cornell University) · Cited by 9 · see citing works
Randomized clinical trials and observational studies in the assessment of drug safety
Sawchik, J., Hamdani, J., & Vanhaeverbeek, M. (2018) · Revue d Épidémiologie et de Santé Publique · doi:10.1016/j.respe.2018.03.133 · Cited by 8 · see citing works
SMITH, R. W., Harty, P. S., Stratton, M. T., Rafi, Z., Rodriguez, C., Dellinger, J. R., Benavides, M. L., Johnson, B. A., White, S. J., Williams, A. D., et al. (2021) · Journal of Functional Morphology and Kinesiology · doi:10.3390/jfmk6020036 · Cited by 11 · see citing works
“ity analysis [38] under the missing not at random (MNAR) assumption using the delta adjustment approach [37] by supposing that individuals who left the study had smaller changes in body composition and exercise strength (Tables S10–S13).”
MethodErdmann, S., Biddle, L., Kieser, M., & Bozorgmehr, K. (2022) · Biometrical Journal · doi:10.1002/bimj.202100045 · Cited by 5 · see citing works
Loilome, W., Dokduang, H., Suksawat, M., & Padthaisong, S. (2021) · Expert Opinion on Investigational Drugs · doi:10.1080/13543784.2021.1955102 · Cited by 6 · see citing works
“powerful method that was used in randomized trial [39].”
MethodIndependence estimators for re-randomisation trials in multi-episode settings: a simulation study
Kahan, B. C., White, I. R., Eldridge, S. M., & Hooper, R. (2021) · BMC Medical Research Methodology · doi:10.1186/s12874-021-01433-4 · Cited by 6 · see citing works
“These results suggest that the re-randomisation design alongside an independence estimator is a potentially useful option to estimate relevant treatment effects in multi-episode settings, though if per-patient or policy-benefit estimators are used it may be useful to conduct sensitivity analyses to evaluate how robust…”
BackgroundSelf-citationCro, S., Morris, T. P., Roger, J., & Carpenter, J. R. (2024) · Pharmaceutical Statistics · doi:10.1002/pst.2373 · Cited by 5 · see citing works
“Reference-based conditional mean imputation cannot be used as a valid sensitivity analysis, because to carry the label sensitivity analysis it must be possible to change inference from the primary analysis; see Morris et al. 9 The p -value- preserving property violates this criterion for J2R and some other choices of…”
Self-citationUddin, M. N., Bashir, N. Z., & Kahan, B. C. (2023) · Clinical Trials · doi:10.1177/17407745231211272 · Cited by 4 · see citing works
Self-citationEede, H. V., Noels, S., Vreven, M., Dujardyn, J., Eede, M. V., Branders, J., Eede, L. V., Herdt, G. D., Noels, I., Bouziotis, J., et al. (2025) · European Journal of Anaesthesiology · doi:10.1097/eja.0000000000002196 · Cited by 4 · see citing works
Hsu, C., He, Y., Hu, C., & Zhou, W. (2023) · Statistics in Medicine · doi:10.1002/sim.9723 · Cited by 4 · see citing works
Kahan, B. C., Diaz‐Ordaz, K., Homer, K., Carnes, D., Underwood, M. R., Taylor, S. J., Bremner, S. A., & Eldridge, S. M. (2014) · Trials · doi:10.1186/1745-6215-15-59 · Cited by 4 · see citing works
“We will perform three sensitivity analyses for the primary outcome to assess the robustness of the results to other methods of account for missing data [32].”
MethodSelf-citationYana, C., & Nainggolan, Y. A. (2023) · International Journal of Current Science Research and Review · doi:10.47191/ijcsrr/v6-i7-54 · Cited by 4 · see citing works
Liossi, C., Georgallis, T., Zhang, J., Hamilton, F., White, P., & Schoth, D. E. (2020) · BMJ Open · doi:10.1136/bmjopen-2019-030607 · Cited by 4 · see citing works
Jose, J., Vishwakarma, G. K., & Bhattacharjee, A. (2021) · Journal of King Saud University - Science · doi:10.1016/j.jksus.2021.101403 · Cited by 4 · see citing works
Jones, C. I., Bremner, S. A., Hooper, R., Gu, J., Bartl, G., & Greenwood, K. (2021) · Trials · doi:10.1186/s13063-021-05670-2 · Cited by 3 · see citing works
Greenwood, K., Jones, C. I., Yaziji, N., Healey, A., May, C. R., Bremner, S. A., Hooper, R., Rathod, S., Phiri, P., de Visser, R. O., et al. (2024) · The British Journal of Psychiatry · doi:10.1192/bjp.2024.154 · Cited by 3 · see citing works
Yao, M., Huan, J., Liang, Q., Jie, W., Wang, M., Ma, Y., Mei, F., Liu, Y., Xiong, Y., Li, L., et al. (2026) · BMC Medical Research Methodology · doi:10.1186/s12874-026-02803-6 · Cited by 1 · see citing works
“Sensitivity analysis is a valuable and widely recommended method for assessing and interpreting the robustness of findings from RCTs, with relevant guidelines available for its implementation [5, 22, 23].”
Shih, W. J. (2016) · Statistics in Medicine · doi:10.1002/sim.6895 · Cited by 1 · see citing works
Gustafson, D. H., Sr., Mares, M., Johnston, D. C., Curtin, J. J., Pe‐Romashko, K., & Landucci, G. (2025) · JMIR Research Protocols · doi:10.2196/64449 · Cited by 1 · see citing works
“Supplemental sensitivity analyses [69] will also be conducted to assess the robustness of conclusions to our assumption that missing data are MAR. Specifically, we will use comparable linear mixed-effects models to examine the effects of the study arm using only complete cases, as would be…”
Greenwood, K., Jones, C. I., Yaziji, N., Healey, A., May, C. R., Bremner, S. A., Hooper, R., Rathod, S., Phiri, P., de Visser, R. O., et al. (2023) · Health and Social Care Delivery Research · doi:10.3310/wdwg4102 · Cited by 1 · see citing works
Sensitivity analyses in pragmatic randomized clinical trials
Suarez, E. A., & Girman, C. J. (2021) · Elsevier eBooks · doi:10.1016/b978-0-12-817663-4.00029-5
The Devil Lies in the Details: An Analysis of Six Oncology Randomised Controlled TrialsPreprint
Dueñas‐González, A., Bornstein-Quevedo, L., & Gutiérrez-Delgado, F. (2025) · Qeios · doi:10.32388/rnymqg
Orlandi, A. (2025) · Journal of Cancer Research and Clinical Oncology · doi:10.1007/s00432-025-06229-3
“Post-hoc analyses often rely on assumptions about missing data patterns that may not reflect reality and typically lack pre-specification, increasing analytical bias (Morris et al. 2014).”
Mwanga, D. M., Kipchirchir, I. C., Muhua, G. O., Newton, C. R. J. C., & Kadengye, D. T. (2025) · Open Journal of Statistics · doi:10.4236/ojs.2025.153018
The Role of Analysis Populations and Estimands in Clinical Trials
Murea, M., Goldman, M. P., Allon, M., & Chen, H. (2025) · Journal of the American Society of Nephrology · doi:10.1681/asn.0000000970
Kahan, B. C., White, I. R., Eldridge, S. M., & Hooper, R. (2021) · Figshare · doi:10.6084/m9.figshare.16909549
Self-citationKahan, B. C., Forbes, G., & Cro, S. (2020) · Figshare · doi:10.6084/m9.figshare.12924162
Self-citationKahan, B. C., White, I. R., Eldridge, S. M., & Hooper, R. (2021) · Figshare · doi:10.6084/m9.figshare.16909546
Self-citationStatistical Analysis Under the Estimand Framework: Continuous Endpoints
Lu, C., Dragalin, V., Polverejan, E., Bornkamp, B., Wolbers, M., Xie, L., & Yang, N. (2025) · Springer series in pharmaceutical statistics · doi:10.1007/978-3-032-02192-2_15
Dopheide, L. H. J., Kozior, M., Buco, A., van Bergenhenegouwen, J., Tims, S., den Bosch, J. D. V., Roberts, J., Jonkers, D. M. A. E., Harvey, L., & van Helvoort, A. (2026) · Current Developments in Nutrition · doi:10.1016/j.cdnut.2026.107695
“Covariate analyses were conducted for primary and secondary outcome parameters, following the sensitivity analysis as out-lined by Morris et al. [36].”
A severity-anchored, patient-reported outcome framework for clinical trials in post-stroke dysphagia
Qureshi, A. I., Kaur, G., Carhuapoma, L., Lakhani, P., Lefton‐Greif, M. A., Mehr, D. R., Royse, L. A., Wang, N.-Y., Zahoor, K., & Hanley, D. F. (2026) · Frontiers in Neurology · doi:10.3389/fneur.2026.1826727
“…alternative severity anchors (e.g., total NIHSS score), and missing-data assumptions (e.g., missing = non-responder; multiple imputation under missing at random (MAR)), with exploratory per-protocol and change-from-baseline analyses where baseline PROs are reliable (62, 63) (Tables 4 , 5).”
- Authors
- Morris, T. P., Kahan, B. C., & White, I. R.
- Affiliations
- MRC Clinical Trials Unit at UCL (GB)
- Queen Mary University of London (GB)
- MRC Biostatistics Unit (GB)
- Corresponding author
- Morris, Tim P
- Venue
- BMC Medical Research Methodology
- Published
- 2014-01-24
- Volume
- 14
- Issue
- 1
- Pages
- 11-11
- ISSN
- 1471-2288 (ISSN-L)
- PMC
- PMC3904008
- OpenAlex
- W2136669583
- Cited by
- 75
- References
- 18
- Citations by year
- 2026: 4 · 2025: 12 · 2024: 6 · 2023: 7 · 2022: 9 · 2021: 9 · 2020: 8 · 2019: 5 · 2018: 3 · 2017: 1 · 2016: 5 · 2015: 3 · 2014: 3
- FWCI
- 4.18
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- BMC Medical Research Methodology (publishedVersion, cc-by)
- Europe PMC (PubMed Central) (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Queen Mary Research Online (Queen Mary University of London) (submittedVersion, cc-by)
- Funders
- Medical Research Council (MC_EX_G0800814, U1052 60558)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Keywords
- sensitivity analysis, randomised trials, primary analysis, trial interpretation
- MeSH terms
- Data Collection, Humans, Research Design, Multicenter Studies as Topic, Randomized Controlled Trials as Topic, Outcome Assessment, Health Care
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.
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
