Common problems related to P-value in observational pharmacoepidemiological studies

doi:10.19803/j.1672-8629.2022.01.05

Abstract

Abstract: Objective To explore the common problems related to P-value in observational pharmacoe-pidemiological studies. Methods A combination of literature research and expert opinions was used for this review. Results The common problems related to P-value in observational pharmacoepidemiological studies included myths, misuse, manipulation and dichotomization of P-value. These problems could compromise the validity and reproducibility of findings and conclusions in pharmacoepidemiological studies. Conclusion Problems related to P-value often occur in observational pharmacoepidemiological studies. An integrated approach that involves advancing researchers' understanding of P-value, enhancing the quality and integrity of research and clinical practices, and ensuring proper use of P-value in the research community will help minimize the misuse of P-value in pharmacoepidemiological studies.

Key words: pharmacoepidemiology, P-value, statistical, observational study

CLC Number:

R994.11

LI Guowei, LI Likang, LIU Yingxin, WANG Ruoting, WU Darong. Common problems related to P-value in observational pharmacoepidemiological studies[J]. Chinese Journal of Pharmacovigilance, 2022, 19(1): 23-26.

References

[1] FISHER RA.Statistical tests[J]. Nature, 1935, 136(3438): 474.
[2] BÉGAUD B. Dictionnary of pharmacoepidemiology[M]. Chichester, UK: John Wiley & Sons Ltd, 2000.
[3] GRIMES DA, SCHULZ KF.Bias and causal associations in observational research[J]. Lancet (London, England), 2002, 359(9302): 248-252.
[4] MA H, LI X, SUN D, et al.Association of habitual glucosamine use with risk of cardiovascular disease: prospective study in UK Biobank[J]. BMJ, 2019, 365: l1628.
[5] WASSERSTEIN RL, LAZAR NA.The ASA statement on p-values: context, process, and purpose[J]. The American Statistician, 2016, 70(2): 129-133.
[6] ALTMAN DG, BLAND JM.Statistics notes: Absence of evidence is not evidence of absence[J]. BMJ, 1995, 311(7003): 485.
[7] SIMMONS JPNL, SIMONSOHN U.False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant[J]. Psychological Science, 2011, 22(11): 1359-1366.
[8] MOTULSKY HJ.Common misconceptions about data analysis and statistics[J]. The Journal of Pharmacology and Experimental Therapeutics, 2014, 351(1): 200-205.
[9] LI G, KAMEL M, JIN Y, et al.Exploring the characteristics, global distribution and reasons for retraction of published articles involving human research participants: a literature survey[J]. Journal of Multidisciplinary Healthcare, 2018, 11: 39-47.
[10] CHAN AW.Bias, spin, and misreporting: time for full access to trial protocols and results[J]. PLoS Medicine, 2008, 5(11): e230.
[11] BOUTRON I, DUTTON S, RAVAUD P, et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes[J]. JAMA, 2010, 303(20): 2058-2064.
[12] WAKEFIELD AJ, MURCH SH, ANTHONY A,et al.Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children[J]. Lancet (London, England), 1998, 351(9103): 637-641.
[13] The Editors of The Lancet. Retracted-ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children[J]. Lancet (London, England), 2010, 375(9713): 445.
[14] LI G, WALTER S, THABANE L.Shifting the focus away from binary thinking of statistical significance and towards education for key stakeholders: revisiting the debate on whether it's time to de-emphasize or get rid of statistical significance[J]. J Clin Epidemiol, 2021, 137: 104-112.
[15] HURLBERT SH, LEVINE RA, UTTS J.Coup de grâce for a tough old bull:“Statistically significant”expires[J]. The American Statistician, 2019, 73(sup1): 352-357.
[16] WASSERSTEIN RL, SCHIRM AL, LAZAR NA.Moving to a world beyond“P< 0.05”[J]. The American Statistician, 2019, 73(sup1): 1-19.
[17] AMRHEIN V, GREENLAND S, MCSHANE B.Scientists rise up against statistical significance[J]. Nature, 2019, 567(7748): 305-307.
[18] BENJAMIN DJ, BERGER JO, JOHANNESSON M, et al.Redefine statistical significance[J]. Nature Human Behaviour, 2018, 2(1): 6-10.
[19] VAN CALSTER B, STEYERBERG EW, COLLINS GS, et al.Consequences of relying on statistical significance: some illustra-tions[J]. European Journal of Clinical Investigation, 2018, 48(5): e12912.
[20] STEYERBERG EW, VAN CALSTER B.Redefining significance and reproducibility for medical research: a plea for higher P-value thresholds for diagnostic and prognostic models[J]. European Journal of Clinical Investigation, 2020, 50(5): e13229.
[21] IOANNIDIS JP.Retiring statistical significance would give bias a free pass[J]. Nature, 2019, 567(7749): 461-462.
[22] FRICKER JR RD, BURKE K, HAN X, et al.Assessing the statistical analyses used in basic and applied social psychology after their P-value ban[J]. The American Statistician, 2019, 73(sup1): 374-384.