As alternatives to the time-to-first-event analysis of composite endpoints, the win statistics, that is, the net benefit, the win ratio, and the win odds have been proposed to assess treatment effects, using a hierarchy of prioritized component outcomes based on clinical relevance or severity. Whether we are using paired organs of a human body or pair-matching patients by risk profiles or propensity scores, we can leverage the level of granularity of matched win statistics to assess the treatment effect. However, inference for the matched win statistics (net benefit, win ratio, and win odds)—quantities related to proportions—is either not available or unsatisfactory, especially in samples of small to moderate size or when the proportion of wins (or losses) is near 0 or 1. In this paper, we present methods to address these limitations. First, we introduce a different statistic to test for the null hypothesis of no treatment effect and provided a sample size formula. Then, we use the method of variance estimates recovery to derive reliable, boundary-respecting confidence intervals for the matched net benefit, win ratio, and win odds. Finally, a simulation study demonstrates the performance of the proposed methods. We illustrate the proposed methods with two data examples.
RedforsBGregsonJCrowleyA, et al. The win ratio approach for composite endpoints: practical guidance based on previous experience. Eur Heart J2020; 41: 4391–4399.
2.
DongGHuangBWangD, et al. Adjusting win statistics for dependent censoring. Pharm Stat2021; 20: 440–450.
3.
MatsouakaRABetenskyRA. Power and sample size calculations for the Wilcoxon–Mann–Whitney test in the presence of death-censored observations. Stat Med2015; 34: 406–431.
4.
PocockSJAritiCACollierTJ, et al. The win ratio: a new approach to the analysis of composite endpoints in clinical trials based on clinical priorities. Eur Heart J2011; 33: 176–182.
5.
BuyseM. Generalized pairwise comparisons of prioritized outcomes in the two-sample problem. Stat Med2010; 29: 3245–3257.
6.
FinkelsteinDMSchoenfeldDA. Combining mortality and longitudinal measures in clinical trials. Stat Med1999; 18: 1341–1354.
7.
DongGHoaglinDCQiuJ, et al. The win ratio: On interpretation and handling of ties. Stat Biopharm Res2020; 12: 99–106.
8.
MilojevicMHeadSJAndrinopoulouER, et al. Hierarchical testing of composite endpoints: applying the win ratio to percutaneous coronary intervention versus coronary artery bypass grafting in the syntax trial. EuroIntervention2017; 13: 106–114.
9.
KotalikAEatonALianQ, et al. A win ratio approach to the re-analysis of multiple risk factor intervention trial. Clin Trials2019; 16: 626–634.
10.
OcchipintiGBucciarelli-DucciCCapodannoD. Diagnostic pathways in myocardial infarction with non-obstructive coronary artery disease (Minoca). Eur Heart J Acute Cardiovasc Care2021; 10: 813–822.
11.
ArmstrongRA. Statistical guidelines for the analysis of data obtained from one or both eyes. Ophthalmic Physiol Opt2013; 33: 7–14.
12.
MurdochIEMorrisSSCousensSN. People and eyes: statistical approaches in ophthalmology. Br J Ophthalmol1998; 82: 971–973.
13.
YingGSMaguireMGGlynnR, et al. Tutorial on biostatistics: statistical analysis for correlated binary eye data. Ophthalmic Epidemiol2018; 25: 1–12.
14.
ListerMTCunninghamMJBenjaminB, et al. Microdebrider tonsillotomy vs electrosurgical tonsillectomy: a randomized, double-blind, paired control study of postoperative pain. Arch Otolaryngol Head Neck Surg2006; 132: 599–604.
15.
BarataTJEBrescianiEMattosMCR, et al. Comparison of two minimally invasive methods on the longevity of glass ionomer cement restorations: short-term results of a pilot study. J Appl Oral Sci2008; 16: 155–160.
16.
BarkerLKGriffinSOJeonS, et al. Ecological-type inference in matched-pair studies with fixed marginal totals. Stat Med2011; 30: 541–548.
17.
ZhuHZhangSAhnC. Sample size considerations for split-mouth design. Stat Methods Med Res2017; 26: 2543–2551.
18.
LesaffreEPhilstromBNeedlemanI, et al. The design and analysis of split-mouth studies: what statisticians and clinicians should know. Stat Med2009; 28: 3470–3482.
19.
ChoYParkJWNamK. The relationship between foot posture index and resting calcaneal stance position in elementary school students. Gait Posture2019; 74: 142–147.
20.
Alfageme-GarcíaPCalderón-GarcíaJFMartínez-NovaA, et al. Backpacks effect on foot posture in schoolchildren with a neutral foot posture: a three-year prospective study. Int J Environ Res Public Health2020; 17: 7313.
21.
Marchena-RodríguezAMoreno-MoralesNRamírez-PargaE, et al. Relationship between foot posture and dental malocclusions in children aged 6 to 9 years: a cross-sectional study. Medicine2018; 97: e0701–e0701.
22.
FuchsNPölzWBathkeAC. Confidence intervals for population means of partially paired observations. Stat Papers2017; 58: 35–51.
23.
MenzHB. Analysis of paired data in physical therapy research: time to stop double-dipping? J Orthop Sports Phys Ther 2005; 35: 477–478.
24.
SuttonAJMuirKRJonesAC. Two knees or one person: data analysis strategies for paired joints or organs. Ann Rheum Dis1997; 56: 401–402.
25.
ZhangYGlynnRFelsonD. Musculoskeletal disease research: should we analyze the joint or the person?. J Rheumatol1996; 23: 1130–1134.
26.
VerbeeckJOzenneBAndersonWN. Evaluation of inferential methods for the net benefit and win ratio statistics. J Biopharm Stat2020; 30: 765–782.
27.
LuoXQiuJBaiS, et al. Weighted win loss approach for analyzing prioritized outcomes. Stat Med2017; 36: 2452–2465.
28.
NewcombeRG. Confidence intervals for proportions and related measures of effect size. London: CRC Press, 2012.
29.
NewcombeRG. Improved confidence intervals for the difference between binomial proportions based on paired data. Stat Med1998; 17: 2635–2650.
30.
ZouGDonnerA. Construction of confidence limits about effect measures: a general approach. Stat Med2008; 27: 1693–1702.
31.
DonnerAZouG. Closed-form confidence intervals for functions of the normal mean and standard deviation. Stat Methods Med Res2012; 21: 347–359.
32.
LachinJM. Biostatistical methods: the assessment of relative risks, vol. 509. New York: John Wiley & Sons, 2009.
33.
AgrestiAMinY. Simple improved confidence intervals for comparing matched proportions. Stat Med2005; 24: 729–740.
34.
BrownLLiX. Confidence intervals for two sample binomial distribution. J Stat Plan Inference2005; 130: 359–375.
CoakleyCWHeiseMA. Versions of the sign test in the presence of ties. Biometrics1996; 52:1242–1251.
37.
BrownLDCaiTTDasGuptaA. Interval estimation for a binomial proportion. Stat Sci2001; 16:101–117.
38.
BrownLDCaiTTDasguptaA. Confidence intervals for a binomial proportion and asymptotic expansions. Ann Stat2002; 30: 160–201.
39.
ZouG. Assessment of risks by predicting counterfactuals. Stat Med2009; 28: 3761–3781.
40.
AgrestiACoullBA. Approximate is better than exact for interval estimation of binomial proportions. Am Stat1998; 52: 119–126.
41.
BebuILachinJM. Large sample inference for a win ratio analysis of a composite outcome based on prioritized components. Biostatistics2015; 17: 178–187.
42.
KatzDBaptistaJAzenS, et al. Obtaining confidence intervals for the risk ratio in cohort studies. Biometrics1978; 34: 469–474.
43.
NewcombeRG. Confidence intervals for an effect size measure based on the Mann–Whitney statistic. Part 2: asymptotic methods and evaluation. Stat Med2006; 25: 559–573.
44.
LuiKJ. Interval estimation of risk ratio in the simple compliance randomized trial. Contemp Clin Trials2007; 28: 120–129.
45.
EfronBTibshiraniRJ. An Introduction to the bootstrap (Chapman & Hall/CRC monographs on statistics and applied probability 57). Boca Raton, FL, USA: chapman & Hall/CRC, 1993.
46.
FiellerEC. Some problems in interval estimation. J R Stat Soc B (Methodol)1954; 16: 175–185.
47.
ZouGDonnerA. A generalization of Fieller’s theorem for ratios of non-normal variables and some practical applications. In: Schuster, H, Metzger, W (eds) Biometrics: Methods, applications and analyses. UK: nova Publications, 2010, pp. 197–216.
48.
DonnerAZouG. Interval estimation for a difference between intraclass kappa statistics. Biometrics2002; 58: 209–215.
49.
R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2020. https://www.R-project.org/
50.
AgrestiA. Categorical data analysis. 3rd ed. Wiley series in probability and statistics. New York: Wiley-Interscience, 2013.
51.
HusterWJBrookmeyerRSelfSG. Modelling paired survival data with covariates. Biometrics1989; 45: 145–156.
52.
LindorKDDicksonERBaldusWP, et al. Ursodeoxycholic acid in the treatment of primary biliary cirrhosis. Gastroenterology1994; 106: 1284–1290.
53.
AgrestiACaffoB. Simple and effective confidence intervals for proportions and differences of proportions result from adding two successes and two failures. Am Stat2000; 54: 280–288.
54.
OzenneBBudtz-JørgensenEPéronJ. The asymptotic distribution of the net benefit estimator in presence of right-censoring. Stat Methods Med Res 2021; 30: 2399–2412.
55.
NewcombeRG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med1998; 17: 857–872.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
