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 601–700 of 1,018
Incidence and Predictors of Atrial Fibrillation in Cardiac Sarcoidosis
Niemelä, M., Uusitalo, V., Pöyhönen, P., Schildt, J., Lehtonen, J. Y. A., & Kupari, M. M. (2022) · JACC. Cardiovascular imaging · doi:10.1016/j.jcmg.2022.02.025 · Cited by 42
Witsell, D. L., Garrett, C. G., Yarbrough, W. G., Dorrestein, S. P., Drake, A. F., & Weissler, M. C. (1995) · PubMed · Cited by 39
Variables are valuable: making a case for deductive modeling
Couto, D. T., & Lorenz, D. (2021) · Linguistics · doi:10.1515/ling-2019-0050 · Cited by 45
Multivariable regression: understanding one of medicine’s most fundamental statistical tools
Varady, N. H., Pareek, A., Eckhardt, C. M., Williams, R. J., Madjarova, S. J., Ollivier, M. P., Martin, R. K., KARLSSON, J. L., & Nwachukwu, B. U. (2022) · Knee Surgery Sports Traumatology Arthroscopy · doi:10.1007/s00167-022-07215-9 · Cited by 39
“There is extensive research and debate about the appropriate EPV threshold, though traditional rules advocate for a minimum of ten EPVs [10–12].”
BackgroundMachine learning for surrogate process models of bioproduction pathways
Huntington, T. T., Baral, N. R., Yang, M., Sundstrom, E. R., & Scown, C. D. (2022) · Bioresource Technology · doi:10.1016/j.biortech.2022.128528 · Cited by 43
Adamuz, J., Viasus, D., CAMPRECIÓS‐RODRÍGUEZ, P., Jurado, O. C., Jiménez‐Martínez, E., Isla, P., García‐Vidal, C., & Carratalà, J. (2011) · Respirology · doi:10.1111/j.1440-1843.2011.02017.x · Cited by 38
Valente, C., Hinds, J., Pinto, F. R., Brugger, S. D., Gould, K. A., Mühlemann, K., de LENCASTRE, H., & Sá‐Leão, R. (2012) · PLoS ONE · doi:10.1371/journal.pone.0030235 · Cited by 46
“Furthermore, according to the later work of Vittinghoff and McCulloch [33], this number of predictors can safely be double if the aim of the analysis is to detect associations in observational data and not making predictions; or if the high number of predictors is necessary to control possible confounding variables.”
BackgroundSchindler, M., Bernard, L., Belaieff, W., Gamulin, A., Racloz, G., Emonet, S., Lew, D., Hoffmeyer, P. J., & Uçkay, İ. (2013) · Journal of Infection · doi:10.1016/j.jinf.2013.07.017 · Cited by 38
First Sexual Intercourse and Subsequent Regret in Three Developing Countries
Osorio, A., del Burgo, C. L., Carlos, S., Ruiz‐Canela, M., Delgado‐Rodríguez, M., & de Irala, J. (2011) · Journal of Adolescent Health · doi:10.1016/j.jadohealth.2011.07.012 · Cited by 46
Li, L. T., Brahmbhatt, R., Hicks, S. C., Davila, J. A., Berger, D. H., & Liang, M. K. (2014) · Digestive Surgery · doi:10.1159/000354426 · Cited by 38
Dedic, A., Genders, T. S. S., Ferket, B. S., Galema, T. W., Mollet, N. R., Moelker, A. D., Hunink, M. G. M., de Feyter, P. J., & Nieman, K. (2011) · Radiology · doi:10.1148/radiol.11110744 · Cited by 40
Loretto, M., Reimann, S., Schuster, R. H., Graulich, D. M., & Bugnyar, T. (2015) · Journal für Ornithologie · doi:10.1007/s10336-015-1289-z · Cited by 41
“To perform the marking, the birds were caught in the game park using three drop-in traps built after Stiehl (1978) without a compartment for decoy birds; instead, we only baited with meat and bread.”
MethodAdenoma Detection Rates Decline with Increasing Procedural Hours in an Endoscopist’s Workload
Almadi, M. A., Sewitch, M. J., Barkun, A. N., Martel, M., & Joseph, L. (2015) · Canadian Journal of Gastroenterology and Hepatology · doi:10.1155/2015/789038 · Cited by 41
“Using the rule of 10 outcome events per predictor variable (20), and the authors’ desire to include up to 12 variables in the multivariable model, it was estimated that 400 screening colonoscopies would be needed to provide sufficient accuracy.”
BackgroundKindler, J., Schultze‐Lutter, F., Michel, C., Martz-Irngartinger, A., Linder, C., Schmidt, S. J., Stegmayer, K., SCHIMMELMANN, B. G., & Walther, S. (2016) · Schizophrenia Research · doi:10.1016/j.schres.2016.04.032 · Cited by 42
for the REDISSEC-CARESS/CCR group, Quintana, J. M., González, N., Antón-Ladislao, A., Redondo, M., Baré, M., de Larrea, N. F., Briones, E., Escobar, A., Sarasqueta, C., et al. (2016) · BMC Cancer · doi:10.1186/s12885-016-2475-y · Cited by 39
“independent variable included in the multivariate logistic regression model [31, 32].”
MethodSender, A., Rutishauser, L., & Staffelbach, B. (2018) · Human Resource Management Journal · doi:10.1111/1748-8583.12183 · Cited by 58
Durojaiye, O. C., Kritsotakis, E. I., Johnston, P. I., Kenny, T., Ntziora, F., & Cartwright, K. E. (2018) · Clinical Microbiology and Infection · doi:10.1016/j.cmi.2018.11.009 · Cited by 38
Invernizzi, M., Michelotti, A., Noale, M., Lopez, G., Runza, L., Giroda, M., Despini, L., Blundo, C., Maggi, S., Gambini, D., et al. (2019) · Journal of Clinical Medicine · doi:10.3390/jcm8020138 · Cited by 38
“The proportional hazard assumption was verified considering Schoenfeld’s residuals of each covariate [36].”
MethodRand, J. S., Fisher, G., Lamb, K., & Hayward, A. (2019) · Frontiers in Veterinary Science · doi:10.3389/fvets.2018.00290 · Cited by 40
“The sample size of both logistic regressions adhered to the established rule of thumb that regression or cox analyses require a minimum of 10 observations per predictor (57), or in the case of a binary logistic regression, a minimum of 5–9 observations per predictor (58).”
MethodNica, A., Kim, S. R., Gien, L. T., Covens, A. L., Bernardini, M. Q., Bouchard‐Fortier, G., Kupets, R., May, T., Vicus, D., Laframboise, S., et al. (2020) · International Journal of Gynecological Cancer · doi:10.1136/ijgc-2020-001816 · Cited by 38
Advanced Statistics: Multiple Logistic Regression, Cox Proportional Hazards, and Propensity Scores
Cioci, A. C., Cioci, A., Mantero, A. M. A., Parreco, J. P., Yeh, D. D., & Rattan, R. (2021) · Surgical Infections · doi:10.1089/sur.2020.425 · Cited by 40
“The conventionally accepted rule of thumb states that there should be a minimum 10 EPV, and no fewer than 20 EPV for risk prediction [15,16].”
van Oort, L., van den Berg, T., Koes, B. W., de Vet, R. H. C. W., Anema, H. J. R., Heymans, M. W., & Verhagen, A. P. (2012) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2012.05.007 · Cited by 43
Levine, A. C., Glavis‐Bloom, J., Modi, P., Nasrin, S., Rege, S., Chu, C., Schmid, C. H., & Alam, N. H. (2015) · Global Health Science and Practice · doi:10.9745/ghsp-d-15-00097 · Cited by 40
“While a minimum number of 5–10 events per variable is required for the derivation of a stable model, the 4 prior studies each had 1–2 events per variable, making it unlikely for them to perform similarly in future datasets.(17,18,22) Indeed, the Clinical Dehydration Scale, the only 1 of these 4 scales to be externally…”
MethodResultLandscape-scale habitat selection by fishers translocated to the Olympic Peninsula of Washington
Lewis, J. C., Jenkins, K. J., Happe, P. J., Manson, D. J., & McCalmon, M. (2016) · Forest Ecology and Management · doi:10.1016/j.foreco.2016.02.032 · Cited by 37
Gavali, H., Mani, K., Furebring, M., Olsson, K. W., Lindstrom, D. J., Sörelius, K., Sigvant, B., Gidlund, K. D., Torstensson, G., Andersson, M., et al. (2021) · European Journal of Vascular and Endovascular Surgery · doi:10.1016/j.ejvs.2021.09.033 · Cited by 37
Betakova, D., Boomgaarden, H. G., Lecheler, S. K., & Schäfer, S. (2024) · Mass Communication & Society · doi:10.1080/15205436.2024.2304759 · Cited by 37
Bauer, L., Langer, R., Becker, K., Hapfelmeier, A., Ott, K., Novotny, A., Höfler, H., & Keller, G. (2012) · PLoS ONE · doi:10.1371/journal.pone.0044566 · Cited by 43
“This sample size and number of events are sufficiently large for a consistent estimation of all effect sizes investigated in this explorative study [23,24].”
BackgroundRaterman, H. G., van Eijk, I. C., Voskuyl, A. E., Peters, M. J. L., Dijkmans, B. A. C., van Halm, V. P., Şimşek, S., Lems, W. F., & Nurmohamed, M. T. (2008) · Annals of the Rheumatic Diseases · doi:10.1136/ard.2008.100776 · Cited by 38
“tests, when appropriate. With RA patients as the reference category, logistic regression analyses were performed to calculate prevalence odds ratios regarding the MetS for hypothyroid RA patients. Five separate models were used: (1) univariable analysis, using the crude, uncorrected data; (2) multivariable analysis, a…”
MethodTateuchi, H., Akiyama, H., Goto, K., So, K., Kuroda, Y., & Ichihashi, N. (2017) · Osteoarthritis and Cartilage · doi:10.1016/j.joca.2017.12.005 · Cited by 40
Bryce, S., Lee, S. J., Ponsford, J. L., Lawrence, R. J., Tan, E. J., & Rossell, S. L. (2018) · Schizophrenia Research · doi:10.1016/j.schres.2018.06.028 · Cited by 42
Barr, P. J., Bonasia, K., Verma, K., Dannenberg, M. D., Yi, C., Andrews, E., Palm, M., Cavanaugh, K. L., Masel, M. C., & Durand, M. (2018) · Journal of Medical Internet Research · doi:10.2196/11308 · Cited by 40
Geeson, C., Wei, L., & Franklin, B. D. (2019) · BMJ Quality & Safety · doi:10.1136/bmjqs-2018-008335 · Cited by 44
Association of Fitness and Grip Strength With Heart Failure
Sillars, A., Celis-Morales, C. A., Ho, F. K., Petermann, F., Welsh, P., Iliodromiti, S., Ferguson, L. D., Lyall, D. M., Anderson, J., Mackay, D. F., et al. (2019) · Mayo Clinic Proceedings · doi:10.1016/j.mayocp.2019.04.041 · Cited by 37
Polovina, M. M., Lund, L. H., Đikić, D., Petrović‐Đorđević, I., Krljanać, G., Milinković, I., Veljić, I., Piepoli, M. F., Rosano, G. M. C., Ristić, A. D., et al. (2019) · European Journal of Heart Failure · doi:10.1002/ejhf.1666 · Cited by 39
Predicting the Risk of Adverse Events in Pregnant Women With Congenital Heart Disease
Chu, R., Chen, W., Song, G., Yao, S., Xie, L., Song, L., Zhang, Y., Chen, L., Zhang, X., Ma, Y., et al. (2020) · Journal of the American Heart Association · doi:10.1161/jaha.120.016371 · Cited by 41
“29 In our study, the number of events per variable is 5.9 in maternal model development cohort and 31 in neonatal model development cohort.”
InfluentialSocial Frailty and Health-Related Quality of Life in Community-Dwelling Older Adults
Ko, Y., & Lee, K. (2022) · International Journal of Environmental Research and Public Health · doi:10.3390/ijerph19095659 · Cited by 43
“For several models of regression, such as multiple regression, a minimum of 10 to 15 observations per predictor variable will generally allow good estimates [25,26].”
MethodHafezi‐Nejad, N., Zikria, B. A., Eng, J., Carrino, J. A., & Demehri, S. (2015) · Skeletal Radiology · doi:10.1007/s00256-015-2217-2 · Cited by 45
Seghrouchni, K., van Delden, C., Dominguez, D. E., Benkabouche, M., Bernard, L., Assal, M., Hoffmeyer, P. J., & Uçkay, İ. (2011) · International Orthopaedics · doi:10.1007/s00264-011-1366-8 · Cited by 37
“We included five to ten predictor variables per outcome event [31].”
GENETIC STUDY: 5‐HTTLPR polymorphism, mood disorders and MDMA use in a 3‐year follow‐up study
Martín‐Santos, R., Torrens, M., Poudevida, S., Langohr, K., Cuyàs, E., Pacifici, R., Farré, M., Pichini, S., & de la Torre, R. (2009) · Addiction Biology · doi:10.1111/j.1369-1600.2009.00180.x · Cited by 48
“Following the general recommendations on the maximum number of variables in regression models, these models were not adjusted for further variables (Vittinghoff & McCulloch 2007) nor were these analyses carried out among the two other study groups.”
Tan, S. K. R., Kolmodin, L., Sekhon, M. S., Qiao, L., Zou, J., Henderson, W. R., & Griesdale, D. E. G. (2016) · Canadian Journal of Anesthesia/Journal canadien d anesthésie · doi:10.1007/s12630-016-0633-y · Cited by 38
Alcohol abstinence and mortality in a general population sample of adults in Germany: a cohort study
John, U., Rumpf, H.-J., Hanke, M., & Meyer, C. (2021) · PLOS Medicine · doi:10.1371/journal.pmed.1003819 · Cited by 36
“We safeguarded that the minimum number of outcome events (death cases) per predictor variable was 5 or higher [30].”
BackgroundRamirez, G. A., Covid-19 BioB Study Group, Bozzolo, E. P., Castelli, E., Marinosci, A., Angelillo, P., Damanti, S., Scotti, R., Gobbi, A., Centurioni, C., et al. (2021) · Minerva Medica · doi:10.23736/s0026-4806.20.06952-9 · Cited by 40
Uyoga, S. M., Olupot‐Olupot, P., Connon, R., Kiguli, S., Opoka, R. O., Alaroker, F., Muhindo, R., Macharia, A. W., Dondorp, A. M., Gibb, D. M., et al. (2022) · The Lancet Child & Adolescent Health · doi:10.1016/s2352-4642(22)00153-5 · Cited by 42
“47 In conclusion, data from the current and other recent studies 17,25 are consistent with an earlier model 6 in which we proposed that even under conditions of intense malaria transmission, SCA affords a degree of innate resistance to high-density malaria infections, and thus protects against…”
Rajwa, P., Robesti, D., Chaloupka, M., Zattoni, F., Giesen, A., Huebner, N., Krzywon, A., Miszczyk, M., Moll, M., Stando, R., et al. (2023) · European Urology Oncology · doi:10.1016/j.euo.2023.09.006 · Cited by 36
“Univariate Cox’s proportional hazard regression models were used for investigating the effect of variables on CRPC-FS and rPFS. HRs (95% CIs) for categorical variables were measured when the following criteria were met: at least ten observations (patients) and at least five events per subgroup [23].”
Olsen, F. J., Diederichsen, S. Z., Jørgensen, P. G., Jensen, M. T., Dahl, A., Landler, N. E., Graff, C., Brandes, A., Krieger, D. W., Haugan, K., et al. (2024) · Circulation Cardiovascular Imaging · doi:10.1161/circimaging.123.016197 · Cited by 45
An osteoarthritis pathophysiological continuum revealed by molecular biomarkers
Kraus, V. B., Sun, S., Reed, A., Soderblom, E. J., Moseley, M. A., Zhou, K., Jain, V., Arden, N., & Li, Y. (2024) · Science Advances · doi:10.1126/sciadv.adj6814 · Cited by 36
“The decision to use 16 as the maximum number of peptides was based on the consideration of the average maximum number of variables permissible in the three models using a 5:1 rule [ratio of min(events, non-events) to number of candidate parameters in the model] ( 61 ).”
MethodLi, P., Li, M., & Chen, W. (2025) · Gut Microbes · doi:10.1080/19490976.2025.2489074 · Cited by 48
“27 Additionally, numerous other steps in model construction can affect performance, 28 with each step offering various execution parameters.”
Griesdale, D. E. G., Örtenwall, V., Norena, M., Wong, H., Sekhon, M. S., Kolmodin, L., Henderson, W. R., & Dodek, P. (2014) · Journal of Critical Care · doi:10.1016/j.jcrc.2014.07.026 · Cited by 38
“Because our aim was to control for confounding of the effect of the exposure variable (CPP time index and CPP area index) and not to build a predictive model, we did not limit the number of covariates in the final model [24].”
MethodKasenda, B., von Elm, E. B., You, J., Blümle, A., Tomonaga, Y., Saccilotto, R., Amstutz, A., Bengough, T., Meerpohl, J., Stegert, M., et al. (2012) · BMC Medical Research Methodology · doi:10.1186/1471-2288-12-131 · Cited by 36
Wang, A. Y.-M., Lam, C. W.-K., Chan, I. H. S., Wang, M., Lui, S., & Sanderson, J. E. (2009) · Nephrology Dialysis Transplantation · doi:10.1093/ndt/gfp325 · Cited by 36
“The rule of 10 events per covariate was relaxed in this situation in order to allow adequate control of confounders [26].”
BackgroundsE-cadherin serves as a diagnostic and predictive parameter in prostate cancer patients
Tsaur, I., Thurn, K., Juengel, E., Gust, K. M., Borgmann, H., Mager, R., Bartsch, G. C., Oppermann, E., Ackermann, H., Nelson, K., et al. (2015) · Journal of Experimental & Clinical Cancer Research · doi:10.1186/s13046-015-0161-6 · Cited by 36
“Multivariate analysis with logistic regression for all 8 candidate markers as well as serum PSA was unsuitable because, taking into account the cohort size and ‘the rule of ten’ [17], sufficient stability of the model could not be ensured.”
MethodAlves, B. E., Montalvao, S. A., Aranha, F. J., Siegl, T. F., Souza, C. A. D., Lorand-Metze, I. G. H., Annichino‐Bizzacchi, J. M., & Paula, E. V. D. (2010) · BMC Infectious Diseases · doi:10.1186/1471-2334-10-143 · Cited by 45
“On the other hand, it has been demonstrated that relaxing the rule of 10 events per variable in Cox models is possible without unacceptable increases in confounding bias [32].”
BackgroundMarconi, D. G., Fregnani, J. H. T. G., Rossini, R. R., Netto, A. K. B. J., Lucchesi, F. R., Tsunoda, A. T., & Kamrava, M. R. (2016) · BMC Cancer · doi:10.1186/s12885-016-2619-0 · Cited by 36
“Also, we limited the number of adjusted variables in the confirmatory multivariate model in order to avoid an over fitting due to the relative small number of events [31].”
MethodCattoni, M., Vallières, E., Brown, L. M., Sarkeshik, A. A., Margaritora, S., Siciliani, A., Filosso, P. L., Guerrera, F., Imperatori, A. S., Rotolo, N., et al. (2017) · Journal of Thoracic and Cardiovascular Surgery · doi:10.1016/j.jtcvs.2017.08.102 · Cited by 36
“The changes in hazard ratio estimates for histology and pT with and without pN in the model were small and not clinically meaningful.(6) We used the results of the multivariable analysis to create survival and recurrence tables on the basis of tumor histology and pathologic tumor stage (pT).”
MethodMa, J., He, Y., Wang, A., Wang, W., Xi, Y., Yu, J., & Ye, X. (2017) · World Neurosurgery · doi:10.1016/j.wneu.2017.11.154 · Cited by 39
Clinical and non-clinical factors that predict discharge disposition after a fall
James, M. K., Robitsek, R. J., Saghir, S. M., Gentile, P. A., Ramos, M., & Perez, F. (2018) · Injury · doi:10.1016/j.injury.2018.02.014 · Cited by 37
Wong, Y., Cheung, C. Y. Y., Tang, C. S.-M., Au, K., Hai, J. S. H., Lee, C., Lau, K. K., Cheung, B. M. Y., Sham, P. C., Xu, A., et al. (2018) · Arteriosclerosis Thrombosis and Vascular Biology · doi:10.1161/atvbaha.118.311726 · Cited by 36
Gilpin, H. R., Stahl, D. R., & McCracken, L. M. (2019) · European Journal of Pain · doi:10.1002/ejp.1310 · Cited by 36
Pavlovic, A. S., Polovina, M. M., Ristić, A. D., Seferović, J. P., Veljić, I., Simeunović, D. S., Milinković, I., Krljanać, G., Ašanin, M., Ostric, I., et al. (2018) · European Journal of Preventive Cardiology · doi:10.1177/2047487318807767 · Cited by 38
Brugger, H., Bouzat, P., Pasquier, M., Mair, P., Fieler, J., Darocha, T., Blancher, M., de Riedmatten, M., Falk, M., Paal, P., et al. (2019) · Resuscitation · doi:10.1016/j.resuscitation.2019.04.025 · Cited by 36
Prevalence of 17 feline behavioral problems and relevant factors of each behavior in Japan
Yamada, R., Arata, S., Kiyokawa, Y., & Takeuchi, Y. (2020) · Journal of Veterinary Medical Science · doi:10.1292/jvms.19-0519 · Cited by 49
“more than 35 cats, which was sufficient to satisfy the requirement based on the “effect of 212 the number of events per variable (EPV)” [32].”
Dual-Plane versus Prepectoral Breast Reconstruction in High–Body Mass Index Patients
Gabriel, A., Sigalove, S. R., Storm-Dickerson, T. L., Sigalove, N. M., Pope, N., Rice, J., & Maxwell, G. P. (2020) · Plastic & Reconstructive Surgery · doi:10.1097/prs.0000000000006840 · Cited by 36
Jiang, C., Kong, Z., Zhang, Y., Liu, S., Liu, Z., Chen, W., Liu, P., Liu, D., Wang, Y., Lyu, Y., et al. (2020) · Neuroradiology · doi:10.1007/s00234-020-02392-1 · Cited by 35
Risk Factors and Survival Analysis of Spinal Cord Stimulator Explantation
Dougherty, M. C., Woodroffe, R. W., Wilson, S., Gillies, G. T., Howard, M. A., & Carnahan, R. M. (2020) · Neuromodulation Technology at the Neural Interface · doi:10.1111/ner.13173 · Cited by 36
“Although bias away from the null related to model overfitting is possible, prior work suggests the use of three common covariates to predict 11 events with Cox regression has low risk for bias (18).”
MethodZhang, S., Wu, X., Bian, S., Zhang, Q., Liu, L., Meng, G., Yao, Z., Wu, H., Gu, Y., Wang, Y., et al. (2020) · British Journal Of Nutrition · doi:10.1017/s0007114520003190 · Cited by 40
“Based on the prevalence of NAFLD in the Chinese population and on the principle of ten outcome events per variable, the sample size was calculated (21) .”
Recurrence risk factors in stage IA grade 1 endometrial cancer
Nwachukwu, C., Baskovic, M., von Eyben, R., Fujimoto, D. K., Giaretta, S., English, D. P.-G., & Kidd, E. A. (2021) · Journal of Gynecologic Oncology · doi:10.3802/jgo.2021.32.e22 · Cited by 42
“Multivariate analysis was performed with all factors significant in a univariate analysis [10].”
MethodGhosn, M., Attallah, N. M., Badr, M., Abdallah, K., Oliveira, B. D., Nadeem, A., Varghese, Y., Munde, D., Salam, S., Abduljawad, B. M., et al. (2021) · Journal of Clinical Medicine · doi:10.3390/jcm10061217 · Cited by 37
“As sensitivity analyses, we evaluated if severe AKI was associated with hospital mortality in different models with less than 10 events per variable [26] by keeping age, severity score, and comorbidity variables in each model.”
BackgroundMethodInfluentialMortality after electroconvulsive therapy
Watts, B. V., Peltzman, T., & Shiner, B. R. (2021) · The British Journal of Psychiatry · doi:10.1192/bjp.2021.63 · Cited by 37
Tortorella, G. L., Fogliatto, F. S., Kurnia, S., Thürer, M., & Capurro, D. (2022) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2022.121780 · Cited by 47
Risk Factors for Complications Differ Between Stages of Tissue-Expander Breast Reconstruction
Lovecchio, F. C., Jordan, S. W., Lim, S., Fine, N. A., & Kim, J. Y. S. (2014) · Annals of Plastic Surgery · doi:10.1097/sap.0000000000000109 · Cited by 35
“20 and an event occurrence of at least 10 for the stage I complication model and at least 5 for the stage II complication model were included as covariates in the multivariate model.(19,20) Hosmer-Lemeshow and C-statistics were computed to assess goodness-of-fit and discriminatory capacity of the model.”
MethodPredicting the Individual Risk of Acute Severe Colitis at Diagnosis
Cesarini, M., Collins, P. G. S., Rönnblom, A., Santos, A., Wang, L. M., Sjöberg, D., Parkes, M., Keshav, S., & Travis, S. P. L. (2016) · Journal of Crohn s and Colitis · doi:10.1093/ecco-jcc/jjw159 · Cited by 38
On the use of propensity scores in case of rare exposure
Hajage, D., Tubach, F., Steg, P. G., Bhatt, D. L., & Rycke, Y. D. (2016) · BMC Medical Research Methodology · doi:10.1186/s12874-016-0135-1 · Cited by 48
“A classical rule when fitting a logistic model is to have an adequate number of outcomes per predictor (at least five or ten outcomes per predictor [49, 50]).”
BackgroundPredictors of dementia in Parkinson disease: A prospective cohort study
Kawada, T. (2015) · Neurology · doi:10.1212/wnl.0000000000001408 · Cited by 36
“Mirror therapy was first used in the treatment of phantom limb pain.(4) The principle of mirror therapy is based on the mirror neuron system.”
MethodRelationships, love and sexuality: what the Filipino teens think and feel
de Irala, J., Osorio, A., del Burgo, C. L., Belen, V. A., de Guzman, F. O., Calatrava, M. D. C., & Torralba, A. N. (2009) · BMC Public Health · doi:10.1186/1471-2458-9-282 · Cited by 35
Simonetti, A. F., Viasus, D., García‐Vidal, C., Adamuz, J., Roset, Á., Manresa, F., Dorca, J., Gudiol, F., & Carratalà, J. (2011) · Clinical Microbiology and Infection · doi:10.1111/j.1469-0691.2011.03709.x · Cited by 35
“We restricted the number of variables included in the multivariable models following the rule of at least five to nine events (deaths) per variable [31].”
MethodBible College Participation and Prison Misconduct: A Preliminary Analysis
Duwe, G., Hallett, M. A., Hays, J., Jang, S. J., & Johnson, B. R. (2015) · Journal of Offender Rehabilitation · doi:10.1080/10509674.2015.1043481 · Cited by 44
“To avoid biased estimates, unreliable confidence interval coverage, and convergence problems in logistic regression models, prior research recommends a rule of thumb of 5 to 10 events per variable (EPV; Penduzzi, Cocato, Kemper, Holford, & Feinstein, 1996; Vittinghoff & McCulloch, 2007).”
MethodA latent variable model approach to estimating systematic bias in the oversampling method
Hauner, K. K. Y., Zinbarg, R. E., & Revelle, W. (2013) · Behavior Research Methods · doi:10.3758/s13428-013-0402-6 · Cited by 35
Outcome prediction in chronic unilateral lumbar radiculopathy: prospective cohort study
Iversen, T., Solberg, T. K., Wilsgaard, T., Waterloo, K. K., Brox, J. I., & Ingebrigtsen, T. (2015) · BMC Musculoskeletal Disorders · doi:10.1186/s12891-015-0474-9 · Cited by 42
“Our study is limited by a relatively small number of patients, which precluded explorative analysis of the effect of different combinations of predictors [53,54].”
León, S. R., Segura, E. R., Konda, K. A., Flores, J. A., Silva‐Santisteban, A., Galea, J. T., Coates, T. J., Klausner, J. D., & Cáceres, C. F. (2016) · BMJ Open · doi:10.1136/bmjopen-2015-008245 · Cited by 42
“In general, to avoid overfitting and instability in the final models, we limited the number of variables to include in the models in such a way that there are about 10 to 16 events in the outcome per dichotomous or dummy.(30) We assessed potential collinearity among predictors by measuring uncentred variance inflation…”
MethodAnalgesics During Pregnancy and Undescended Testis
Philippat, C., Giorgis-Allemand, L., Chevrier, C., Cordier, S., Jégou, B., Charles, M.-A. A., & Slama, R. (2011) · Epidemiology · doi:10.1097/ede.0b013e318225bf33 · Cited by 35
Penn, A. M. W., Bibok, M. B., Saly, V. K., Coutts, S. B., Lesperance, M., Balshaw, R. F., Votova, K., Croteau, N. S., Trivedi, A., Jackson, A., et al. (2018) · Biomarkers · doi:10.1080/1354750x.2018.1434681 · Cited by 35
Hendrickx, L. A. M., Cain, M., Sierevelt, I. N., Jadav, B., Kerkhoffs, G. M. M. J., Jaarsma, R. L., & Doornberg, J. N. (2019) · Journal of Orthopaedic Trauma · doi:10.1097/bot.0000000000001605 · Cited by 35
Amoah, P. A. (2019) · International Journal of Environmental Research and Public Health · doi:10.3390/ijerph16173188 · Cited by 47
“Nonetheless, as argued above, the sample was adequate in view of the nature of the models constructed [64,65].”
BackgroundPredicting Concussion Outcome by Integrating Finite Element Modeling and Network Analysis
Anderson, E. D., Giudice, J. S., Wu, T., Panzer, M. B., & Meaney, D. F. (2020) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2020.00309 · Cited by 39
“Additionally, because we had a limited number of cases, we were not able to leverage a large number of potential concussion predictors at once and were limited to up to four at a time (Vittinghoff and McCulloch, 2006).”
BackgroundMethodAnalysis of factors affecting the prognosis of COVID-19 patients and viral shedding duration
Han, J., Shi, L., Xie, Y. F., Zhang, Y.-J., Huang, S.-P., Li, J., Wang, H.-R., & Shao, S.-F. (2020) · Epidemiology and Infection · doi:10.1017/s0950268820001399 · Cited by 36
Roberts, D. J., Nagpal, S., Kubelik, D., Brandys, T., Stelfox, H. T., Lalu, M. M., Forster, A. J., McCartney, C. J. L., & McIsaac, D. I. (2020) · BMJ · doi:10.1136/bmj.m4104 · Cited by 45
Tangri, N., Reaven, N. L., Funk, S. E., Ferguson, T. W., Collister, D., & Mathur, V. S. (2021) · BMC Nephrology · doi:10.1186/s12882-021-02385-z · Cited by 38
“Previous studies examining the association between serum bicarbonate levels and CKD progression have found conflicting results, potentially due to differences in sample sizes, patient populations, baseline risks of end-stage kidney disease, and duration of follow-up as well as the definition and number of events, the…”
ResultLee, J. W., Philteos, J., Levin, M., Namavarian, A., Propst, E. J., & Wolter, N. E. (2021) · JAMA Otolaryngology–Head & Neck Surgery · doi:10.1001/jamaoto.2021.1548 · Cited by 42
Ramos, L. A., van Os, H. J. A., Hilbert, A., Olabarriaga, S. D., van der Lugt, A., Roos, Y. B. W. E. M., van Zwam, W. H., van Walderveen, M. A. A., Ernst, M. S., Zwinderman, A. H., et al. (2022) · Frontiers in Neurology · doi:10.3389/fneur.2022.809343 · Cited by 42
“CRP level at baseline—median (IQR) 4 (2–10) 651 (20) Cont”
BackgroundDunkel, N., Uçkay, I., Belaieff, W., Assal, M., Corni, V., Lacraz, A., & Karaca, Ş. (2012) · Journal of Orthopaedic Science · doi:10.1007/s00776-012-0245-5 · Cited by 35
“We included 6–10 outcome events per predictor variable [16].”
BackgroundRates and correlates of sexual activity and impairment among women with systemic sclerosis
Levis, B., Hudson, M., Knafo, R., Baron, M. G., Nielson, W. R., Hill, M. L., & Thombs, B. D. (2011) · Arthritis Care & Research · doi:10.1002/acr.20696 · Cited by 40
“Age group, years 18–39 25 13 (52) 7 (54) 7 (28) 40–49 103 67 (65) 39 (58) 39 (38) 50–59 171 81 (47) 27 (33) 27 (16) 60 248 65 (26) 22 (34) 22 (9) Marital status* Married 373 189 (51) 75 (40) 75 (20) Not married 167 33 (20) 18 (55) 18 (11) Total 547 226 (41) 95 (42) 95 (17)”
InfluentialYalcin, S., & Leroux, S. (2018) · Global Change Biology · doi:10.1111/gcb.14169 · Cited by 37
Effects of age and sex on clinical high-risk for psychosis in the community
Schultze-Lutter, F., Schimmelmann, B. G., Flückiger, R., & Michel, C. (2020) · World Journal of Psychiatry · doi:10.5498/wjp.v10.i5.101 · Cited by 34
Sane, J., MacDonald, E. A., Vold, L., Gossner, C. M., Severi, E., & collective on behalf of the International Outbreak Investigation Team (2015) · Eurosurveillance · doi:10.2807/1560-7917.es2015.20.4.21018 · Cited by 34
“25, excluding those with fewer than 10 cases [10].”
BackgroundFactors Associated With Treatment Failure of Infected Pressure Sores
Jugun, K., Richard, J.-C., Lipsky, B. A., Kressmann, B., Pittet-Cuenod, B. M., Suvà, D., Modarressi, A., & Uçkay, İ. (2015) · Annals of Surgery · doi:10.1097/sla.0000000000001497 · Cited by 34
“We included 5 to 8 predictor variables per outcome event.(10) We checked key variables for colinearity, and for interaction (by Mantel–Haenszel estimates and interaction terms).”
Impact of a high loading dose of amikacin in patients with severe sepsis or septic shock
Allou, N., Bouteau, A., Allyn, J., Snauwaert, A., Valance, D., Jabot, J., Bouchet, B., Galliot, R., Corradi, L., Montravers, P., et al. (2016) · Annals of Intensive Care · doi:10.1186/s13613-016-0211-z · Cited by 43
“We considered that 50 patients were needed in the non-survivors group to perform a multivariate analysis with ten variables [13].”
MethodShifting brucellosis risk in livestock coincides with spreading seroprevalence in elk
Brennan, A., Cross, P. C., Portacci, K., Scurlock, B. M., & Edwards, W. H. (2017) · PLoS ONE · doi:10.1371/journal.pone.0178780 · Cited by 42
Eustatic sea-level controls on the flushing of a shelf-incising submarine canyon
Allin, J. R., Hunt, J. E., Clare, M. A., & Talling, P. J. (2017) · Geological Society of America Bulletin · doi:10.1130/b31658.1 · Cited by 34
“Previous work has shown that the Cox PH model requires at least a minimum sample size of N = 20 (Vittinghoff and McCulloch, 2007).”
MethodHevesi, M., Hartigan, D. E., Wu, I. T., Levy, B. A., Domb, B. G., & Krych, A. J. (2018) · The American Journal of Sports Medicine · doi:10.1177/0363546518767399 · Cited by 35
Factors influencing fire suppression success in the province of Quebec (Canada)
Cardíl, A., Lorente, M., Boucher, D., Boucher, J., & Gauthier, S. (2018) · Canadian Journal of Forest Research · doi:10.1139/cjfr-2018-0272 · Cited by 39
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.