Comparison of the Andersen-Gill model with poisson and negative binomial regression on recurrent event data

Many generalizations of the Cox proportional hazard method have been elaborated to analyse recurrent event data. The Andersen-Gill model was proposed to handle event data following Poisson processes. This method is compared with non-survival approaches, such as Poisson and negative binomial regression. The comparison is performed on data simulated according to various event-generating processes and differing in subject heterogeneity. When robust standard error estimates are applied, for Poisson processes the Andersen-Gill approach is comparable to a negative binomial regression, whereas the poisson regression has comparable coverage probabilities of confidence intervals, but increased type I error rates; however, none of the methods can generate unbiased parameter estimates with data violating the independent increment assumption. These findings are illustrated by data from a clinical trial of the efficacy of a new pneumococcal vaccine for prevention of otitis media.

[1]  R J Cook,et al.  Marginal analysis of recurrent events and a terminating event. , 1997, Statistics in medicine.

[2]  P F Thall,et al.  Mixed Poisson likelihood regression models for longitudinal interval count data. , 1988, Biometrics.

[3]  S. Chevret,et al.  Analysis of recurrent failure times data: should the baseline hazard be stratified , 2001, Statistics in medicine.

[4]  Norman E. Breslow,et al.  Tests of Hypotheses in Overdispersed Poisson Regression and other Quasi-Likelihood Models , 1990 .

[5]  A. V. Peterson,et al.  On the regression analysis of multivariate failure time data , 1981 .

[6]  J. Poolman,et al.  Pneumococcal capsular polysaccharides conjugated to protein D for prevention of acute otitis media caused by both Streptococcus pneumoniae and non‐typable Haemophilus influenzae: a randomised double‐blind efficacy study , 2006, Lancet.

[7]  J. Vaupel,et al.  The impact of heterogeneity in individual frailty on the dynamics of mortality , 1979, Demography.

[8]  R. Gill,et al.  Cox's regression model for counting processes: a large sample study : (preprint) , 1982 .

[9]  C. Dean,et al.  Efficiency Lost by Analyzing Counts Rather than Event Times in Poisson and Overdispersed Poisson Regression Models , 1997 .

[10]  Simon G Thompson,et al.  The importance of varying the event generation process in simulation studies of statistical methods for recurrent events , 2006, Statistics in medicine.

[11]  S. Thompson,et al.  Wei, Lin and Weissfeld's marginal analysis of multivariate failure time data , 2007, Statistical methods in medical research.

[12]  L. J. Wei,et al.  Regression analysis of multivariate incomplete failure time data by modeling marginal distributions , 1989 .

[13]  M. Gorfine,et al.  On robustness of marginal regression coefficient estimates and hazard functions in multivariate survival analysis of family data when the frailty distribution is mis‐specified , 2007, Statistics in medicine.

[14]  J. Buring,et al.  Ways of measuring rates of recurrent events , 1996, BMJ.

[15]  R. Glynn,et al.  Estimating the variance of standardized rates of recurrent events, with application to hospitalizations among the elderly in New England. , 1993, American journal of epidemiology.

[16]  L. J. Wei,et al.  The Robust Inference for the Cox Proportional Hazards Model , 1989 .

[17]  P J Kelly,et al.  Survival analysis for recurrent event data: an application to childhood infectious diseases. , 2000, Statistics in medicine.

[18]  T. Stukel,et al.  Comparison of methods for the analysis of longitudinal interval count data. , 1993, Statistics in medicine.

[19]  M. Greenwood,et al.  An Inquiry into the Nature of Frequency Distributions Representative of Multiple Happenings with Particular Reference to the Occurrence of Multiple Attacks of Disease or of Repeated Accidents , 1920 .

[20]  Z Guo,et al.  Modeling Repeated Time-to-event Health Conditions with Discontinuous Risk Intervals , 2008, Methods of Information in Medicine.

[21]  T M Therneau,et al.  rhDNase as an example of recurrent event analysis. , 1997, Statistics in medicine.

[22]  R J Cook,et al.  Robust tests for treatment comparisons based on recurrent event responses. , 1996, Biometrics.

[23]  E. Parner,et al.  Methods for analysing recurrent events in health care data. Examples from admissions in Ebeltoft Health Promotion Project. , 2006, Family practice.