二元成對資料在臨床試驗中相當常見，其中比例差常為研究關注之焦點。實務上常因資料缺失使得分析受限；而在多重插補法框架下，現有處理此類缺失資料之區間估計方法，仍面臨覆蓋率不足之缺陷，且此問題在參數趨近邊界時尤為嚴重。為克服此限制，本研究在多重插補框架下，建構並比較多種改良之區間估計程序。模擬結果顯示，本研究提出之新方法能有效改善覆蓋率偏低的情形，其整體表現最為穩定且優於現有方法。此外，在維持良好覆蓋率的前提下，該程序具備較短的平均區間寬度，能提供更精確的區間估計。最後，本研究亦透過真實資料集展示各項方法之應用價值。

Matched-pair binary data are frequently encountered in clinical trials, where the difference in proportions is often the primary focus of interest. In practice, data missingness often restricts analysis. Within the framework of multiple imputation (MI), existing interval estimation methods for such missing data still suffer from insufficient coverage probabilities, a problem that is particularly pronounced when parameters approach their boundaries. To address these limitations, this study develops and compares several improved interval estimation procedures within the MI framework. Simulation results show that the proposed modified MI-MOVER effectively mitigates the issue of undercoverage, demonstrating the most stable overall performance and outperforming existing methods. Furthermore, while maintaining satisfactory coverage probabilities, the proposed procedure yields shorter average interval widths, thereby providing more precise interval estimates. Finally, the practical value of the proposed method is demonstrated through the analysis of real-world datasets.

[1] Chang, P., Liu, R., Hou, T., Yan, X., and Shan, G., "Continuity corrected score confidence interval for the difference in proportions in paired data," Journal of Applied Statistics, vol. 51, no. 1, 2024.
[2] Das, U. and Basu, R., "Approximate confidence intervals for the difference in proportions for partially observed binary data," Statistical Methods in Medical Research, vol. 31, no. 3, pp. 488-509, 2022.
[3] Harel, O. and Zhou, X. H., "Multiple imputation for the comparison of two screening tests in two‐phase Alzheimer studies," Statistics in Medicine, vol. 26, no. 11, pp. 2370-2388, 2007.
[4] Kadane, J. B., "Is victimization chronic? A Bayesian analysis of multinomial missing data," Journal of econometrics, vol. 29, no. 1-2, pp. 47-67, 1985.
[5] Lee, C. H. and Wang, H., "Multiple imputation confidence intervals for the mean of the discrete distributions for incomplete data," Statistics in Medicine, vol. 41, no. 7, pp. 1172-1190, 2022.
[6] Lee, C. H., "Multiple Imputation Confidence Intervals for a Risk Difference With Missing Observations," Statistics in Medicine, vol. 44, no. 13-14, p. e70136, 2025.
[7] Li, X., Mehrotra, D. V., and Barnard, J., "Analysis of incomplete longitudinal binary data using multiple imputation," Statistics in Medicine, vol. 25, no. 12, pp. 2107-2124, 2006.
[8] Lott, A. and Reiter, J. P., "Wilson confidence intervals for binomial proportions with multiple imputation for missing data," The American Statistician, vol. 74, no. 2, pp. 109-115, 2020.
[9] Newcombe, R. G., "Improved confidence intervals for the difference between binomial proportions based on paired data," Statistics in medicine, vol. 17, no. 22, pp. 2635-2650, 1998.
[10] Osoba, D., Erlichman, C., Willan, A., Levitt, M., and Pater, J., "Superiority of methylprednisolone sodium succinate over low dose metoclopramide hydrochloride in the prevention of nausea and vomiting produced by cancer chemotherapy," Clinical and Investigative medicine. Medecine Clinique et Experimentale, vol. 9, no. 4, pp. 225-231, 1986.
[11] Prentice, R. L., "Correlated binary regression with covariates specific to each binary observation," Biometrics, pp. 1033-1048, 1988.
[12] Rubin, D. B., Multiple Imputation for Nonresponse in Surveys. Wiley, 2004.
[13] Schafer, J., Analysis of Incomplete Multivariate Data. CRC Press, 1997.
[14] Sidi, Y. and Harel, O., "Difference between binomial proportions using Newcombe’s method with multiple imputation for incomplete data," The American Statistician, vol. 76, no. 1, pp. 29-36, 2022.
[15] Tang, N. S., Tang, M. L., and Chan, I. S. F., "On tests of equivalence via non‐unity relative risk for matched‐pair design," Statistics in medicine, vol. 22, no. 8, pp. 1217-1233, 2003.
[16] Tango, T., "Equivalence test and confidence interval for the difference in proportions for the paired‐sample design," Statistics in medicine, vol. 17, no. 8, pp. 891-908, 1998.
[17] Wei, J., He, P., and Tong, T., "Estimating the reciprocal of a binomial proportion," International Statistical Review, vol. 92, no. 1, pp. 1-16, 2024.
[18] Xiu, L., Xie, L., Yan, H., Wu, C., Liu, H., and Chen, C., "MOVER tests for non-inferiority of the difference between two binary-outcome treatments in the matched-pairs design," Journal of Biopharmaceutical Statistics, vol. 36, no. 2, pp. 213-226, 2026.
[19] Yang, Z., Sun, X., and Hardin, J. W., "A non‐iterative implementation of Tango's score confidence interval for a paired difference of proportions," Statistics in Medicine, vol. 32, no. 8, pp. 1336-1342, 2013.
[20] Zhang, S., Cao, J., and Ahn, C., "Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes," Statistics in medicine, vol. 36, no. 4, pp. 581-591, 2017.