Relaxing the Rule of Ten Events per Variable in Logistic and Cox Regression
The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
The rule of thumb that logistic and Cox models should be used with a minimum of 10 outcome events per predictor variable (EPV), based on two simulation studies, may be too conservative. The authors conducted a large simulation study of other influences on confidence interval coverage, type I error, relative bias, and other model performance measures. They found a range of circumstances in which coverage and bias were within acceptable levels despite less than 10 EPV, as well as other factors that were as influential as or more influential than EPV. They conclude that this rule can be relaxed, in particular for sensitivity analyses undertaken to demonstrate adequate control of confounding.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Following the generally accepted rule of thumb that 5–10 events are required per covariate in logit or Cox regression models (Peduzzi et al., 1995; Peduzzi et al., 1996; Vittinghoff & McCulloch, 2007), we assessed whether or not each paper could have supported fully saturated regression models with product terms between all of the psychosocial problems considered.”
From Syndemics of psychosocial problems and HIV risk: A systematic review of empirical tests of the disease interaction concept (2015) · cites this paper as Vittinghoff & McCulloch, 2007
In the citing paper
“Because so few students used psychotherapy, we limited the number of explanatory variables to the number of observed events divided by five (14).”
From Mental health treatment seeking among veteran and civilian community college students (2017) · cites this paper as [14]
In the citing paper
“This sample size was chosen taking into account approximate sample size estimation criteria [38,39]. We worked with the criteria that 10 subjects would be needed per parameter included in a statistical model used to adjust for confounding.”
From Relationships, love and sexuality: what the Filipino teens think and feel (2009) · cites this paper as [38]
In the citing paper
“As Vittinghoff and McCulloch 27 recommended, the 10 outcome events per predictor variable rule was relaxed.”
From Mental health services for parents who lost a child to cancer: if we build them, Will they come? (2015) · cites this paper as [27]
In the citing paper
“The requirements regarding events per variable were fulfilled owing to a considerable number of injuries in the present dataset.34”
From How much running is too much? Identifying high-risk running sessions in a 5200-person cohort study (2025) · cites this paper as [34]
Citations
Cites
A simulation study of the number of events per variable in logistic regression analysis
Peduzzi, P., Concato, J., Kemper, E., Holford, T. R., & Feinstein, A. R. (1996) · Journal of Clinical Epidemiology · doi:10.1016/s0895-4356(96)00236-3 · Cited by 9,655 · cited 1×
“(2, 3) more closely, we also examined models with all binary predictors.”
BackgroundModeling and variable selection in epidemiologic analysis
Greenland, S. (1989) · American Journal of Public Health · doi:10.2105/ajph.79.3.340 · Cited by 2,208
Peduzzi, P., Concato, J., Feinstein, A. R., & Holford, T. R. (1995) · Journal of Clinical Epidemiology · doi:10.1016/0895-4356(95)00048-8 · Cited by 2,154 · cited 2×
“(2, 3) more closely, we also examined models with all binary predictors.”
BackgroundResultThe Abuse of Power: The Pervasive Fallacy of Power Calculations for Data Analysis
Hoenig, J. M., & Heisey, D. M. (2001) · The American Statistician · doi:10.1198/000313001300339897 · Cited by 2,046
Sample-Size Formula for the Proportional-Hazards Regression Model
Schoenfeld, D. A. (1983) · Biometrics · doi:10.2307/2531021 · Cited by 1,096
Concato, J., Peduzzi, P., Holford, T. R., & Feinstein, A. R. (1995) · Journal of Clinical Epidemiology · doi:10.1016/0895-4356(95)00510-2 · Cited by 797 · cited 1×
“The rule of thumb that logistic and Cox models should be used with a minimum of 10 events per predictor variable (EPV) is based on two simulation studies (1–3).”
MethodSection 4 Long-term survival results in medically and surgically randomized patients
Detre, K. M., Peduzzi, P. N., Scott, S. M., & Davies, B. (1985) · Progress in Cardiovascular Diseases · doi:10.1016/0033-0620(85)90019-2 · Cited by 24
Section 2 Design and baseline characteristics
Peduzzi, P. N., Detre, K. M., & Gage, A. A. (1985) · Progress in Cardiovascular Diseases · doi:10.1016/0033-0620(85)90017-9 · Cited by 15
Harrell, F., Jr. (1996) · doi:10.1002/(sici)1097-0258(19960229)15:4<361::aid-sim168>3.0.co;2-4
Cited by
Items 701–800 of 1,018
Brown, E., Tohira, H., Bailey, P. M., Fatovich, D. M., Pereira, G. F., & Finn, J. (2018) · Prehospital Emergency Care · doi:10.1080/10903127.2018.1551451 · Cited by 34
“The cutoff values for the tRTS informed categorization of the physiological parameters and these were determined so that there would be at least 10 patients per category to reduce bias and avoid unreliable confidence interval calculation (21).”
MethodModin, D., Biering-Sørensen, S. R., Møgelvang, R., Jensen, J. S., & Biering‐Sørensen, T. (2018) · Circulation Cardiovascular Imaging · doi:10.1161/circimaging.117.007528 · Cited by 34
Peripheral Nerve Blocks for Ambulatory Shoulder Surgery
Hamilton, G. M., Ramlogan, R. R., Lui, A. C. P., McCartney, C. J. L., Abdallah, F. W., McVicar, J. A., & McIsaac, D. I. (2019) · Anesthesiology · doi:10.1097/aln.0000000000002865 · Cited by 35
Development of a prediction model for inpatient gout flares in people with comorbid gout
Jatuworapruk, K., Grainger, R., Dalbeth, N., & Taylor, W. J. (2019) · Annals of the Rheumatic Diseases · doi:10.1136/annrheumdis-2019-216277 · Cited by 33
Inflammatory pathway analytes predicting rapid cognitive decline in MCI stage of Alzheimer’s disease
Pillai, J. A., Bena, J., Bebek, G., Bekris, L. M., Bonner‐Jackson, A., Kou, L., Pai, A., Sørensen, L., Neilsen, M., Rao, S. M., et al. (2020) · Annals of Clinical and Translational Neurology · doi:10.1002/acn3.51109 · Cited by 36
“The ADmark Alzheimer’s evaluation uses sandwich Enzyme Linked Immunosorbant Assay(ELISA) kits [Innotest b-amyloid[1-42], Innotest hTAU-Ag, Innotest Phospho-Tau[181P], Innogenetics, Ghent, Belgium].”
MethodLuis, B. A. L., Sifuentes‐Osornio, J., Lambraño-Castillo, D., Ortíz‐Brizuela, E., Ramírez-Fontes, A., Tovar-Calderón, Y. E., Leal-Vega, F. J., Bobadilla-del-Valle, M., & Ponce‐de‐León, A. (2020) · Journal of Global Antimicrobial Resistance · doi:10.1016/j.jgar.2020.12.005 · Cited by 37
Clinical Implications of Poststent Optical Coherence Tomographic Findings
Kim, B. G., Kachel, M., Kim, J., Guagliumi, G., Kim, C., Kim, I., Lee, Y., Lee, O., Byun, Y. S., Kim, B. O., et al. (2021) · JACC. Cardiovascular imaging · doi:10.1016/j.jcmg.2021.03.008 · Cited by 34
Brayner, B., Kaur, G., Keske, M. A., Perez‐Cornago, A., Piernas, C., & Livingstone, K. M. (2021) · Journal of Nutrition · doi:10.1093/jn/nxab275 · Cited by 37
“Logistic regression requires at least 10 cases of the outcome in question for each covariate in the study (42, 43).”
MethodJenkins, L. C., Chang, W., Buscemi, V., Liston, M. B., Humburg, P., Nicholas, M. K., Graven‐Nielsen, T., Hodges, P. W., McAuley, J. H., & Schabrun, S. M. (2022) · Pain · doi:10.1097/j.pain.0000000000002684 · Cited by 39
Takeuchi, N., Hiratani, M., & Kawano, R. (2022) · JACS Au · doi:10.1021/jacsau.2c00117 · Cited by 34
“47 For this study, we would therefore need more than 5 × 5 = 25 samples because there were fi ve types of miRNAs.”
MethodChau, N. G., Florescu, A., Chan, K., Wang, L. E., Chen, E. X., Bedard, P. L., Oza, A. M., & Siu, L. L. (2011) · BMC Cancer · doi:10.1186/1471-2407-11-426 · Cited by 34
“A recent simulation study suggests that the commonly used 10:1 rule for the number of predictors per events may be overly stringent and can be relaxed [21].”
MethodBalasubramanya, S., Zurakowski, D., Borisuk, M. J., Kaza, A. K., Emani, S. M., del Nido, P. J., & Baird, C. W. (2017) · Journal of Thoracic and Cardiovascular Surgery · doi:10.1016/j.jtcvs.2017.09.019 · Cited by 38
Outcome Effects on Recidivism Among Drug Court Participants
Gibbs, B. R., Lytle, R. D., & Wakefield, W. (2018) · Criminal Justice and Behavior · doi:10.1177/0093854818800528 · Cited by 35
Ajith, M. M., & Ghosh, A. K. (2019) · Safety Science · doi:10.1016/j.ssci.2019.04.010 · Cited by 33
Prognostic value of left atrioventricular coupling index in heart failure
Kasa, G., Teis, A., Raffele, M. D., Cediel, G., Juncà, G., Lupón, J., Santiago‐Vacas, E., Codina, P., Bayés‐Genís, A., & Delgado, V. (2025) · European Heart Journal - Cardiovascular Imaging · doi:10.1093/ehjci/jeaf010 · Cited by 36
Uçkay, İ., Agostinho, A., Belaieff, W., Toutous‐Trellu, L., Scherer-Pietramaggiori, S. S., Andrés, A., Bernard, L., Vuagnat, H., Hoffmeyer, P. J., & Wyssa, B. (2011) · World Journal of Surgery · doi:10.1007/s00268-011-0993-y · Cited by 33
“The number of variables in the final model was limited to the ratio of 7–10 variables to 1 outcome events [ 27 ].”
BackgroundElens, L., Sombogaard, F., Hesselink, D. A., van Schaik, R. H. N., & van Gelder, T. (2013) · Pharmacogenetics and Genomics · doi:10.1097/fpc.0000000000000001 · Cited by 37
Künzel, J., Mantsopoulos, K., Psychogios, G. V., Agaimy, A., Grundtner, P., Koch, M., & Iro, H. (2014) · European Archives of Oto-Rhino-Laryngology · doi:10.1007/s00405-014-2997-3 · Cited by 33
Serum biomarkers of treatment response within a randomized clinical trial for pulmonary tuberculosis
Jayakumar, A., Vittinghoff, E., Segal, M. R., Kenzie, W. R. M., Johnson, J. L., Gitta, P., Saukkonen, J. J., Anderson, J., WEINER, M. H., Engle, M., et al. (2015) · Tuberculosis · doi:10.1016/j.tube.2015.04.011 · Cited by 40
“Nevertheless, the crossvalidation technique employed here serves to mitigate any potential bias in our results based on the relatively low events per variable and small sample size, which would otherwise be vulnerable to over-fitting[43].”
MethodSelf-citationBertot, L. C., Vilar‐Gómez, E., Gonzalez, A. T., Socias-López, M., Moisés Diago, M., Adams, L. A., & Romero‐Gómez, M. (2015) · Digestive and Liver Disease · doi:10.1016/j.dld.2015.12.002 · Cited by 33
for the IRYSS-COPD group, Esteban, C., Aróstegui, I., García‐Gutiérrez, S., González, N., Lafuente, I., Baré, M., de Larrea, N. F., Rivas‐Ruiz, F., & Quintana, J. M. (2015) · Respiratory Research · doi:10.1186/s12931-015-0313-4 · Cited by 36
“When developing clinical prediction models with a binary outcome, a recommended sample size of 10 events per predictor is an extended rule [29–31].”
MethodPredictors of Infant Age at Enrollment in Early Infant Diagnosis Services in Kenya
Goggin, K., Wexler, C., Nazir, N., Staggs, V. S., Gautney, B. J., Okoth, V., Khamadi, S. A., Ruff, A., Sweat, M. D., Cheng, A.-L., et al. (2016) · AIDS and Behavior · doi:10.1007/s10461-016-1404-z · Cited by 33
“Given the small size of the ‘‘very late’’ group, we took a less conservative approach and set the tuning parameter in the ‘‘very late’’ model to select seven candidate predictors, resulting in five ‘‘very late’’ observations per variable in the final logistic regression model [39].”
BackgroundStructural and Process Factors That Influence Clinical Nurse Specialist Role Implementation
Kilpatrick, K., Tchouaket, E., Carter, N., Bryant-Lukosius, D., & DiCenso, A. (2016) · Clinical Nurse Specialist · doi:10.1097/nur.0000000000000182 · Cited by 32
“Given that no previous work has been undertaken in this area, researchers recommend using 10 to 20 cases per variable.(43,44) We estimated the sample size to detect a large effect size (0.”
Absence of Fitness Improvement Is Associated with Outcomes in Heart Failure Patients
Bakker, E. A., Snoek, J. A., Meindersma, E. P., Hopman, M. T. E., Bellersen, L., Verbeek, A. L. M., Thijssen, D. H. J., & Eijsvogels, T. M. H. (2017) · Medicine & Science in Sports & Exercise · doi:10.1249/mss.0000000000001429 · Cited by 40
“To develop a prediction model, we used the rule of thumb with minimal 10 events per variable for obtaining a good prediction (14,15).”
MethodFukazawa, K., Pretto, E. A., Nishida, S., Reyes, J. D., & Gologorsky, E. (2018) · Journal of Clinical Anesthesia · doi:10.1016/j.jclinane.2017.10.017 · Cited by 33
Predicting early cognitive decline in newly-diagnosed Parkinson's patients: A practical model
Hogue, O., Fernandez, H. H., & Floden, D. P. (2018) · Parkinsonism & Related Disorders · doi:10.1016/j.parkreldis.2018.06.031 · Cited by 39
“We elected to relax the ten events per variable convention in order to adequately control for the effect of participant education on the outcome [13].”
MethodThe spatial pattern of redwood burl poaching and implications for prevention
Kurland, J., Pires, S. F., & Marteache, N. (2018) · Forest Policy and Economics · doi:10.1016/j.forpol.2018.06.009 · Cited by 35
Wainer, Z., Wright, G. M., Gough, K., Daniels, M. G., Russell, P. A., Choong, P., Conron, M., Ball, D., & Solomon, B. (2018) · Clinical Lung Cancer · doi:10.1016/j.cllc.2018.08.004 · Cited by 37
Forcino, R. C., Yen, R. W., Aboumrad, M., Barr, P. J., Schubbe, D. C., Elwyn, G. J., & Durand, M. (2018) · BMJ Open · doi:10.1136/bmjopen-2018-022730 · Cited by 37
“31 32 Procedure To advertise the survey, SERMO distributed email invita-tions to members of its survey panel; see online supplementary appendix A2 for the email invitation.”
Estimation of Association between Healthcare System Efficiency and Policy Factors for Public Health
Lee, S., & Kim, C. (2018) · Applied Sciences · doi:10.3390/app8122674 · Cited by 41
“Considering that this restriction can be slightly relaxed in most cases [35], we decided to select up three independent variables.”
BackgroundMortality after inpatient treatment for diarrhea in children: a cohort study
Talbert, A. W., Ngari, M. M., Bauni, E., Mwangome, M., Mturi, N., Otiende, M., Maitland, K., Walson, J. L., & Berkley, J. A. (2019) · BMC Medicine · doi:10.1186/s12916-019-1258-0 · Cited by 37
Casey, J. R., Banks, J. L., Braniff, K., Buettner, P. G., & Heal, C. F. (2019) · Australian and New Zealand Journal of Obstetrics and Gynaecology · doi:10.1111/ajo.12966 · Cited by 34
Crespo, R. D. J., Wu, J., Myer, M. H., Yee, S. H., & Fulford, R. S. (2019) · The Science of The Total Environment · doi:10.1016/j.scitotenv.2019.04.287 · Cited by 33
Gass, J. S., Mitchell, S. D., & Hanna, M. (2019) · BMC Cancer · doi:10.1186/s12885-019-5553-0 · Cited by 37
Microsatellite Instability and Adjuvant Chemotherapy in Stage II Colon Cancer
Koenig, J. L., Toesca, D. A. S., Harris, J. P., Tsai, C. J., Haraldsdóttir, S., Lin, A. Y., Pollom, E. L., & Chang, D. T. (2019) · American Journal of Clinical Oncology · doi:10.1097/coc.0000000000000554 · Cited by 36
Factors associated with surgeon’s perception of distraction in the operating room
Jung, J. J., Elfassy, J., & Grantcharov, T. (2019) · Surgical Endoscopy · doi:10.1007/s00464-019-07088-z · Cited by 33
Petricciuolo, S., Donne, M. G. D., Aimo, A., Chella, A., & Caterina, R. D. (2020) · European Journal of Clinical Investigation · doi:10.1111/eci.13400 · Cited by 37
Rohmann, J. L., Huo, S., Sperber, P. S., Piper, S. K., Rosendaal, F. R., Heuschmann, P. U., Endres, M., Liman, T., & Siegerink, B. (2020) · Journal of Thrombosis and Haemostasis · doi:10.1111/jth.15092 · Cited by 38
Barker, K. L., Room, J., Knight, R., Dutton, S., Toyé, F., Leal, J., Kent, S., Kenealy, N., Schussel, M. M., Collins, G. S., et al. (2020) · Health Technology Assessment · doi:10.3310/hta24650 · Cited by 36
The Perception of Flood Risks: A Case Study of Babessi in Rural Cameroon
Buchenrieder, G. R., Brandl, J., & Balgah, A. R. (2021) · International Journal of Disaster Risk Science · doi:10.1007/s13753-021-00345-7 · Cited by 40
Xu, J. J., Reijnierse, E. M., Pacifico, J., Wan, C. S., & Maier, A. B. (2021) · Age and Ageing · doi:10.1093/ageing/afab134 · Cited by 34
“A modified EWGSOP algorithm was used in the latter study [34].”
BackgroundMethodAdvancing cervical cancer diagnosis and screening with spectroscopy and machine learning
Ramírez, C., Greenop, M., Almoshawah, Y. A., Martin‐Hirsch, P. L., & Rehman, I. U. (2023) · Expert Review of Molecular Diagnostics · doi:10.1080/14737159.2023.2203816 · Cited by 45
“A common rule of thumb is to use 10 observations per feature, resulting in a 1000 wavenumber spectrum requiring the collection of 10,000 spectra if the entire spectrum was included (although this is not always the case [45].”
BackgroundTracking Social Motivation Systems Deficits: The Affective Neuroscience View of Autism
Carré, A., Chevallier, C., Robel, L., Barry, C., Maria, A., Pouga, L., Philippe, A., Pinabel, F., & Berthoz, S. (2015) · Journal of Autism and Developmental Disorders · doi:10.1007/s10803-015-2498-2 · Cited by 43
“…rule of thumb is that logistic models should be used with a minimum of five or even 10 outcome events per predictor variable (Peduzzi et al. 1996; Vittinghoff and McCulloch 2007), and given the present sample size (20 subjects per group), we did not design a single model accounting for all…”
MethodGiacchetti, S., Porcher, R., Lehmann‐Che, J., Hamy, A., de Roquancourt, A., Cuvier, C., Cottu, P. H., Bertheau, P., Albiter, M., Bouhidel, F. A., et al. (2014) · British Journal of Cancer · doi:10.1038/bjc.2014.81 · Cited by 35
“Although we decided not to conform to the ‘rule of thumb’ of 10 events per variable (Vittinghoff and McCulloch, 2007), we kept the number of predictors as low as possible.”
MethodGlucose tolerance before and after renal transplantation
Bergrem, H. A., Valderhaug, T. G., Hartmann, A., Bergrem, H., Hjelmesaeth, J. S., & Jenssen, T. G. (2009) · Nephrology Dialysis Transplantation · doi:10.1093/ndt/gfp566 · Cited by 35
“This produced >5 events per variable (EPV), which has been shown to produce similar results as compared to the commonly employed but probably overly strict 10 EPV rule of thumb [20].”
ResultPersonality, Emotions and Coping Styles: Predictive Value for the Evolution of Cancer Patients
Cardenal, V., Cerezo, M. V., Martínez, J., Ortiz‐Tallo, M., & Blanca, M. J. (2012) · The Spanish Journal of Psychology · doi:10.5209/rev_sjop.2012.v15.n2.38887 · Cited by 42
“There is a ratio of 16 subjects per predictor, which satisfies the rule of thumb of using a minimum of 10 subjects per predictor (Peduzzi, Cocanto, & Kemper, 1996), although recent findings recommend relaxing this rule (Vittinghoff & McCulloch, 2007).”
BackgroundThe Impact of Caregiving on Caregivers’ Medication Adherence and Appointment Keeping
Wang, X., Robinson, K. M., & Hardin, H. K. (2014) · Western Journal of Nursing Research · doi:10.1177/0193945914533158 · Cited by 40
“The total number of variables examined in this study was nine (four demographic variables plus five care demand factors), and thus, 45 caregivers were recruited for this small pilot study based on a minimum rule of 5 to 10 participants per predictor (Vittinghoff & McCulloch, 2007).”
MethodA preliminary study on how hypohydration affects pain perception
Bear, T., Philipp, M. C., Hill, S., & Mündel, T. (2016) · Psychophysiology · doi:10.1111/psyp.12610 · Cited by 40
Anxiety and Adverse Coronary Artery Disease Outcomes in Chinese Patients
Wang, G., Cui, J., Wang, Y., Biyong, D., Liang, X., Bai, J., Guo, S., Yang, Z., Huang, L., & Li, C. (2013) · Psychosomatic Medicine · doi:10.1097/psy.0b013e3182984317 · Cited by 35
“Vittinghoff E, McCulloch CE.”
BackgroundIllness Beliefs Predict Mortality in Patients with Diabetic Foot Ulcers
Vedhara, K., Dawe, K., Miles, J. N. V., Wetherell, M. A., Cullum, N. A., Dayan, C. M., Drake, N., Price, P., Tarlton, J. F., Weinman, J. A., et al. (2016) · PLoS ONE · doi:10.1371/journal.pone.0153315 · Cited by 36
“Although this resulted in our models having up to ten predictors, this approach is in keeping with contexts in which it is appropriate to relax the rule of ten predictors per number of outcomes[22]; As both the predictor and outcome variables contained missing values, imputation methods were used to maximise the avail…”
MethodWatanabe, H., Morimoto, T., Shiomi, H., Furukawa, Y., Nakagawa, Y., Andò, K., Kadota, K., & Kimura, T. (2017) · EuroIntervention · doi:10.4244/eij-d-15-00421 · Cited by 33
Ville, S., Imbert‐Marcille, B., Coste‐Burel, M., Garandeau, C., Meurette, A., Cantarovich, D., Giral, M., Hourmant, M., Blancho, G., & Dantal, J. (2017) · Transplant International · doi:10.1111/tri.13085 · Cited by 35
“Even if others found that this number could be relaxed until 5–9 [38], given the number of EPV in our study (10/4 = 2.”
BackgroundSchmidt, B. T., Reinhard, M., Lezaic, V., McLeod, D. D., Weinhold, M., Mattes, H., & Klingelhöfer, J. (2015) · Journal of Clinical Monitoring and Computing · doi:10.1007/s10877-015-9726-3 · Cited by 33
“Therefore, even if relaxing the commonly used rule of ten events per independent variable [29], we could not consider more than two”
MethodJinks, R. C., Royston, P., & Parmar, M. K. (2015) · BMC Medical Research Methodology · doi:10.1186/s12874-015-0078-y · Cited by 32
“A later simulation study found that in ‘a range of circumstances’ having less than 10 EPV still provided acceptable confidence interval coverage and bias when using Cox regression, but did not directly consider the statistical power of analyses nor the variability of the estimates [14].”
MethodPloegstra, M., Ivy, D. D., Beghetti, M., Bonnet, D., Alehan, D., Ablonczy, L., Mattos, S. D. S., Bowers, D., Humpl, T., Berger, R. M. F., et al. (2023) · European Heart Journal - Quality of Care and Clinical Outcomes · doi:10.1093/ehjqcco/qcad020 · Cited by 42
“19 In the long-term outcome models, all adverse outcome events during the full follow-up period were considered: Patients who did not undergo an adverse outcome event were censored at the last recorded visit.”
BackgroundMachine learning‐based prediction models for pressure injury: A systematic review and meta‐analysis
Pei, J., Guo, X., Tao, H., Wei, Y., Zhang, H., Ma, Y., & Han, L. (2023) · International Wound Journal · doi:10.1111/iwj.14280 · Cited by 41
Sperotto, F., Lang, N., Nathan, M., Kaza, A. K., Hoganson, D. M., Valencia, E., Odegard, K. C., Allan, C. K., da Cruz, E. M., del Nido, P. J., et al. (2023) · Circulation Cardiovascular Interventions · doi:10.1161/circinterventions.123.013383 · Cited by 32
Traeger, A. C., Henschke, N., Hübscher, M., Williams, C. M., Kamper, S. J., Maher, C. G., Moseley, G. L., & McAuley, J. H. (2015) · BMJ Open · doi:10.1136/bmjopen-2015-007916 · Cited by 38
Risk Factors for Delayed Inpatient Functional Recovery after Total Knee Arthroplasty
Hoogeboom, T. J., van Meeteren, N. L. U., Schank, K., Kim, R. H., Miner, T. M., & Stevens‐Lapsley, J. E. (2015) · BioMed Research International · doi:10.1155/2015/167643 · Cited by 37
“This ratio was greater than 5 (the lowest acceptable limit) [17], indicating that the models were sufficiently powered.”
BackgroundGram and acridine orange staining for diagnosis of septic arthritis in different patient populations
Cunningham, G. J., Seghrouchni, K., Ruffieux, E., Vaudaux, P. E., Gayet‐Ageron, A., Cherkaoui, A., Godinho, E., Lew, D., Hoffmeyer, P. J., & Uçkay, İ. (2014) · International Orthopaedics · doi:10.1007/s00264-014-2284-3 · Cited by 32
Flaqué, M. C. D., Galigniana, N. M., Béguelin, W., Vicario, R., Proietti, C. J., Russo, R. I. C., Rivas, M. A., Tkach, M., Guzmán, P., Roa, J. C., et al. (2013) · Breast Cancer Research · doi:10.1186/bcr3587 · Cited by 32
“R Adjustment for significant confounders was done to avoid increased bias and variability or unreliable confidence interval coverage [42].”
MethodRetractedEffect of Inspiratory Muscle Training Before Cardiac Surgery in Routine Care
Valkenet, K., de Heer, F., BACKX, F. J. G., Trappenburg, J. C. A., Hulzebos, E. H. J., Kwant, S., van Herwerden, L. A., & van de Port, I. G. L. (2013) · Physical Therapy · doi:10.2522/ptj.20110475 · Cited by 33
Radtke, J., Dietze, N., Spetzler, V. N., Fischer, L., Achilles, E. G., Li, J., Scheidat, S., Thaiss, F., Nashan, B., & Koch, M. (2015) · Transplant Infectious Disease · doi:10.1111/tid.12494 · Cited by 32
“As each predictor variable needs a minimum number of outcome events, prior variable selection for multivariate models was conducted (27).”
Method25-Hydroxyvitamin-D and Bone Turnover Marker Levels in Patients with Distal Radial Fracture
Rozental, T. D., Herder, L. M., Walley, K. C., Zurakowski, D., Coyle, K., Bouxsein, M., & Wolf, J. M. (2015) · Journal of Bone and Joint Surgery · doi:10.2106/jbjs.o.00313 · Cited by 35
The effects of landscape history and time-lags on plant invasion in Mediterranean coastal habitats
González‐Moreno, P., Pino, J., Cózar, A., García‐de‐Lomas, J., & Vilà, M. (2016) · Biological Invasions · doi:10.1007/s10530-016-1314-z · Cited by 33
“This threshold is at the lower limit of the 30–300 plots recommended for an average binomial model of 6 independent variables (Peduzzi et al. 1996; Vittinghoff and McCulloch 2007).”
Huibertse, L. J., van Eenbergen, M. C. H. J., de Rooij, B. H., Bastiaens, M. T., Fossion, L. M. C. L., de la Fuente, R. B., Kil, P. J. M., Koldewijn, E. L., Meier, A. H. P., Mommers, R. J. M., et al. (2017) · Acta Oncologica · doi:10.1080/0284186x.2016.1267398 · Cited by 32
“variables was allowed in the multinomial logistic regression analysis [28].”
MethodImplementation of Neurocritical Care Is Associated With Improved Outcomes in Traumatic Brain Injury
Sekhon, M. S., Gooderham, P. A., Toyota, B. D., Kherzi, N., Hu, V., Dhingra, V. K., Hameed, M. S., Chittock, D. R., & Griesdale, D. E. G. (2017) · Canadian Journal of Neurological Sciences / Journal Canadien des Sciences Neurologiques · doi:10.1017/cjn.2017.25 · Cited by 37
“Jugular bulb use, n (%) 17 (15) 9 (12) 8 (22) 0.”
Nguyen, T., Collins, G. S., Spence, J. D., Devereaux, P. J., Daurès, J., Landais, P., & Manach, Y. L. (2017) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.4325 · Cited by 37
“As the PS has reductive dimension properties, it removes the problem of the number of events per variable in logistic regression.(12,13)”
BackgroundMethodUithoven, K. E., Smith, J. R., Medina‐Inojosa, J. R., Squires, R. W., & Olson, T. P. (2020) · Journal of Cardiac Failure · doi:10.1016/j.cardfail.2020.01.011 · Cited by 38
Gu, S., Zhou, J., Yuan, C., & Ye, Q. (2020) · Journal of Affective Disorders · doi:10.1016/j.jad.2020.02.046 · Cited by 37
ICU discharge screening for prediction of new‐onset physical disability—A multinational cohort study
Milton, A., Schandl, A. R., Soliman, I., Joelsson‐Alm, E., van den Boogaard, M., Wallin, E., Brorsson, C., Östberg, U., Latocha, K. M., Savilampi, J., et al. (2020) · Acta Anaesthesiologica Scandinavica · doi:10.1111/aas.13563 · Cited by 38
“Duration of mechanical ventilation, hours, median (IQR) 46 (15-188) 42 (7-119) .”
BackgroundInfluentialLee, J. W., Kim, S. Y., Han, S. W., Lee, J. E., Lee, H. J., Heo, N. H., & Lee, S. M. (2020) · EJNMMI Research · doi:10.1186/s13550-020-00660-y · Cited by 32
Gait speed and adverse outcomes following hospitalised exacerbation of COPD
Walsh, J. A., Barker, R. E., Kon, S. S. C., Jones, S. E., BANYA, W. A. S., Nolan, C. M., Patel, S., Polgar, O., Haselden, B. M., Polkey, M. I., et al. (2021) · European Respiratory Journal · doi:10.1183/13993003.04047-2020 · Cited by 35
van Dijk, L. V., & Fuller, C. D. (2021) · American Society of Clinical Oncology Educational Book · doi:10.1200/edbk_320951 · Cited by 44
“76 For deep learning, upsampling methods have been proposed to overcome the shortcoming of imbalanced data; however, similar to general machine learning, the training performance measures must be evaluated with caution, and effectiveness needs to be better investigated 77 ; regardless, the model…”
MethodGonzalez‐Estrada, A., Campbell, R. L., Carrillo‐Martin, I., Renew, J. R., Rank, M. A., & Volcheck, G. W. (2021) · British Journal of Anaesthesia · doi:10.1016/j.bja.2021.06.036 · Cited by 36
Indoor Air Pollution and Susceptibility to Tuberculosis Infection in Urban Vietnamese Children
Blount, R. J., Phan, H., Trinh, T., Dang, H., Merrifield, C., Zavala, M., Zabner, J., Comellas, A. P., Stapleton, E. M., Segal, M. R., et al. (2021) · American Journal of Respiratory and Critical Care Medicine · doi:10.1164/rccm.202101-0136oc · Cited by 41
Oyama, T., Kinoshita, H., Takekawa, D., Saito, J., Kushikata, T., & Hirota, K. (2021) · Journal of Anesthesia · doi:10.1007/s00540-021-03007-6 · Cited by 32
“However, a recent simulation study suggested that 5–9 events per predictor variable is sufficient [15].”
BackgroundFactors for severe outcomes following SARS-CoV-2 infection in people with cystic fibrosis in Europe
Jung, A., Orenti, A., Dunlevy, F., Aleksejeva, E., Bakkeheim, E., Bobrovnichy, V., Carr, S. B., Colombo, C., Corvol, H., Cosgriff, R., et al. (2021) · ERJ Open Research · doi:10.1183/23120541.00411-2021 · Cited by 34
Tohidinezhad, F., Bontempi, D., Zhang, Z., Dingemans, A. C., Aerts, J. G. J. V., Bootsma, G. P., Vansteenkiste, J., Hashemi, S. M. S., Smit, E. F., Gietema, H. A., et al. (2023) · European Journal of Cancer · doi:10.1016/j.ejca.2023.01.027 · Cited by 34
Koch, M., Kantas, A., Ramcke, K., Drabik, A. I., & Nashan, B. (2015) · Clinical Transplantation · doi:10.1111/ctr.12513 · Cited by 32
“Due to the limited number of events, it was not possible to include all of the risk factors or combinations of risk factors in the multivariate analyses (8).”
BackgroundMethodInfluentialCriteria used when deciding on eligibility for total knee arthroplasty — Between thinking and doing
Skou, S. T., Roos, E. M., Laursen, M. B., Rathleff, M. S., Arendt‐Nielsen, L., Simonsen, O. H., & Rasmussen, S. (2015) · The Knee · doi:10.1016/j.knee.2015.08.012 · Cited by 32
“6 The 200 patients allowed us to conduct the analyses using a minimum of 20 predictor variables 120 [20].”
MethodVittinghoff, E., Glidden, D. V., SHIBOSKI, S. C., & McCulloch, C. E. (2012) · Statistics in the health sciences · doi:10.1007/978-1-4614-1353-0_10 · Cited by 36
Self-citationMassachusetts’ parental consent law and procedural timing among adolescents undergoing abortion
Janiak, E., Fulcher, I. R., Cottrill, A. A., Tantoco, N., Mason, A. H., Fortin, J., Sabino, J., & Goldberg, A. B. (2019) · Obstetrics & Gynecology · doi:10.1097/aog.0000000000003190 · Cited by 31
Dessap, A. M., Papazian, L., Schaller, M., Nseir, S., Mégarbane, B., Haudebourg, L., Timsit, J., Teboul, J., Kuteifan, K., Gainnier, M., et al. (2023) · Annals of Intensive Care · doi:10.1186/s13613-023-01150-9 · Cited by 32
“[37] suggest that this rule of thumb can be relaxed.”
BackgroundTrajectories of pain and opioid use up to one year after surgery: analysis of a European registry
Hofer, D. M., Harnik, M. A., Lehmann, T., Stüber, F., Baumbach, P., Dreiling, J., Meißner, W., & Stamer, U. M. (2024) · British Journal of Anaesthesia · doi:10.1016/j.bja.2023.12.002 · Cited by 35
Serdoz, L. V., Daleffe, E., Merlo, M., Zecchin, M., Barbati, G., Pecora, D., Pinamonti, B., Fantoni, C., Lupo, P., Lenarda, A. D., et al. (2011) · The American Journal of Cardiology · doi:10.1016/j.amjcard.2011.02.347 · Cited by 31
Metzger, A., Wakschlag, L. S., Anderson, R., Darfler, A., Price, J., Flores, Z., & Mermelstein, R. J. (2013) · Developmental Psychology · doi:10.1037/a0030720 · Cited by 33
“…per predictor variable (Peduzzi, Concato, Kemper, Holford, & Feinstein, 1996), but recent research has found that this rule can be relaxed under some circumstances, specifically for smaller sample observational studies hoping to control for confounding variables (Vittinghoff & McCulloch, 2007).”
Nakafero, G., Sanders, R., Nguyen‐Van‐Tam, J. S., & Myles, P. R. (2016) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.4028 · Cited by 32
Bua, N., Smith, G., Totty, J. P., Pan, D., Wallace, T., Carradice, D., & Chetter, I. C. (2017) · Annals of Vascular Surgery · doi:10.1016/j.avsg.2017.03.198 · Cited by 33
Improved estimation of the cumulative incidence of rare outcomes
Benkeser, D. C., Carone, M., & Gilbert, P. B. (2017) · Statistics in Medicine · doi:10.1002/sim.7337 · Cited by 36
“However, these approaches similarly suffer from erratic behavior in the context of rare events [10].”
BackgroundKoenig, J. L., Shi, S., Sborov, K. D., Gensheimer, M. F., Li, G., Nagpal, S., Chang, S. D., Gibbs, I. C., Soltys, S. G., & Pollom, E. L. (2019) · World Neurosurgery · doi:10.1016/j.wneu.2019.03.110 · Cited by 33
van der Meer, E., Botman, S., & Eckhardt, S. (2019) · PLoS ONE · doi:10.1371/journal.pone.0215211 · Cited by 38
Arnalich‐Montiel, F., Mingo‐Botín, D., & de-Arriba-Palomero, P. (2019) · American Journal of Ophthalmology · doi:10.1016/j.ajo.2019.07.012 · Cited by 32
Dias, A. R., Fernandes, S. M., Silva, I. T. F., Silva, T. C., Miranda‐Scippa, Â., & Almeida, A. G.-D. (2021) · Trends in Psychiatry and Psychotherapy · doi:10.47626/2237-6089-2020-0187 · Cited by 56
“Median (IQR) 22 (21-25) 23 (21-26) 23 (21-26) 0.”
Egner, L. E., Klöckner, C. A., & Pellegrini‐Masini, G. (2021) · Energy and Buildings · doi:10.1016/j.enbuild.2021.111542 · Cited by 32
Bernardo, R. J., Lu, D., Ramírez, R. L., Hedlin, H., Kawut, S. M., Bull, T., Marco, T. D., Ford, H. J., Grinnan, D., Klinger, J. R., et al. (2022) · Annals of the American Thoracic Society · doi:10.1513/annalsats.202109-1051oc · Cited by 34
Sardi, J., Lázaro, B. C. R., Smith, J. S., Kelly, M. P., Dial, B. L., Hills, J. M., Yanik, E. L., Gupta, M. C., Baldus, C. R., Yen, C. P., et al. (2022) · Journal of Neurosurgery Spine · doi:10.3171/2022.8.spine22423 · Cited by 35
Chung, V. C., Ma, H., Wang, H., Wang, J. J., Hong, L. C., Wei, X., Wong, S. Y. S., Tang, J. L., & Griffiths, S. M. (2013) · Evidence-based Complementary and Alternative Medicine · doi:10.1155/2013/426360 · Cited by 31
“logistic regression analyses, we have more than 10 utilization events per independent variable; thus our sample size is sufficient for conducting such analyses [21].”
MethodHolley, A. B., Petteys, S., Mitchell, J. D., Holley, P. R., & Collen, J. F. (2013) · CHEST Journal · doi:10.1378/chest.12-2879 · Cited by 32
Brown, M. M., Vaclavik, D., Watson, D. P., & Wilka, E. (2017) · Psychological Services · doi:10.1037/ser0000112 · Cited by 31
“It has been suggested that five or more events (i.e., re-entries) per variable is adequate for multivariate Cox proportional hazards analyses (Vittinghoff & McCulloch, 2007).”
Method
Metadata
- Authors
- Vittinghoff, E., & McCulloch, C. E.
- Affiliations
- University of California, San Francisco (US)
- Venue
- American Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2007-01-12
- Volume
- 165
- Issue
- 6
- Pages
- 710-718
- ISSN
- 0002-9262 (ISSN-L)
- 1476-6256
- OpenAlex
- W2130373985
- Cited by
- 4,359
- References
- 9
- Citations by year
- 2026: 642 · 2025: 394 · 2024: 313 · 2023: 315 · 2022: 306 · 2021: 351 · 2020: 332 · 2019: 281 · 2018: 282 · 2017: 205 · 2016: 194 · 2015: 171 · 2014: 146 · 2013: 132 · 2012: 94
- FWCI
- 20.12
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- American Journal of Epidemiology (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Keywords
- sample size, logistic regression, Cox proportional hazards model, confidence interval coverage, type I error, relative bias, simulation study, sensitivity analysis, confounding adjustment, small sample size
- MeSH terms
- Data Interpretation, Statistical, Factor Analysis, Statistical, Humans, Predictive Value of Tests, Sensitivity and Specificity, Bias, Confounding Factors, Epidemiologic, Prevalence, Multivariate Analysis, Confidence Intervals, Logistic Models, Proportional Hazards Models, Case-Control Studies, Sample Size
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
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.