Influence functions for proportional hazards regression

Reid, N.; Crépeau, Hélène

doi:10.1093/biomet/72.1.1

Cited by 87 publications

(31 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To analyze the risk factors of V. parahaemolyticus diarrhea or cholera, we used survival analyses of the time to the first episode of the disease, censoring the follow-up of individuals who died or migrated out [22]. We fitted unadjusted and covariate adjusted Cox proportional hazard regression models verifying first that the proportionality assumption was satisfied for all independent variables [23-25]. In both covariate-unadjusted and covariate-adjusted analyses of the risk for V. parahaemolyticus diarrhea or cholera, we accounted for the design effect induced by cluster randomization by use of robust sandwich variance estimates [25].…”

Section: Methodsmentioning

confidence: 99%

Clinical, epidemiological, and spatial characteristics of Vibrio parahaemolyticus diarrhea and cholera in the urban slums of Kolkata, India

et al. 2012

View full text Add to dashboard Cite

BackgroundThere is not much information on the differences in clinical, epidemiological and spatial characteristics of diarrhea due to V. cholerae and V. parahaemolyticus from non-coastal areas. We investigated the differences in clinical, epidemiological and spatial characteristics of the two Vibrio species in the urban slums of Kolkata, India.MethodsThe data of a cluster randomized cholera vaccine trial were used. We restricted the analysis to clusters assigned to placebo. Survival analysis of the time to the first episode was used to analyze risk factors for V. parahaemolyticus diarrhea or cholera. A spatial scan test was used to identify high risk areas for cholera and for V. parahaemolyticus diarrhea.ResultsIn total, 54,519 people from the placebo clusters were assembled. The incidence of cholera (1.30/1000/year) was significantly higher than that of V. parahaemolyticus diarrhea (0.63/1000/year). Cholera incidence was inversely related to age, whereas the risk of V. parahaemolyticus diarrhea was age-independent. The seasonality of diarrhea due to the two Vibrio species was similar. Cholera was distinguished by a higher frequency of severe dehydration, and V. parahaemolyticus diarrhea was by abdominal pain. Hindus and those who live in household not using boiled or treated water were more likely to have V. parahaemolyticus diarrhea. Young age, low socioeconomic status, and living closer to a project healthcare facility were associated with an increased risk for cholera. The high risk area for cholera differed from the high risk area for V. parahaemolyticus diarrhea.ConclusionWe report coexistence of the two vibrios in the slums of Kolkata. The two etiologies of diarrhea had a similar seasonality but had distinguishing clinical features. The risk factors and the high risk areas for the two diseases differ from one another suggesting different modes of transmission of these two pathogens.

show abstract

Section: Methodsmentioning

confidence: 99%

Clinical, epidemiological, and spatial characteristics of Vibrio parahaemolyticus diarrhea and cholera in the urban slums of Kolkata, India

et al. 2012

View full text Add to dashboard Cite

show abstract

“…The statistic A(m -1~k is the diagnostic given by Cain and Lange (1984), equation 2, and Reid and Crepeau's (1985) empirical influence function is identical with this statistic.…”

Section: Jeri Ijerimentioning

confidence: 98%

“…For standard least squares regression, the sample influence curve can be computed exactly, but for the proportional hazards model~(k) is found by iterative numerical techniques. Various approximations have been suggested for the proportional hazards model: the case deletion location outlier model of Storer and Crowley (1985), the infinitesimal jackknife approach of Cain and Lange (1984) and the influence function approach of Reid and Crepeau (1985). The first uses a one-step Newton-Raphson approximation while the last two give the same statistic.…”

Section: Case Influence Measuresmentioning

confidence: 99%

Case-Weighted Measures of Influence for Proportional Hazards Regression

Pettitt¹,

Daud²

1989

Applied Statistics

View full text Add to dashboard Cite

show abstract

“…It has been well documented in the literature that the Cox model is sensitive even to slight departures from the assumptions (Samuels (1978), Bednarski (1989), Minder and Bednarski (1996)), and that its IF is not bounded (Reid and Crépeau, 1985). Valsecchi et al (1996) provide a detailed illustration of how long survivors, for instance, may affect the estimates.…”

Section: Robust Survival Analysismentioning

confidence: 99%

An overview of robust methods in medical research

Farcomeni

Ventura

2010

Stat Methods Med Res

View full text Add to dashboard Cite

Robust statistics is an extension of classical parametric statistics that specifically takes into account the fact that the assumed parametric models used by the researchers are only approximate. In this paper we review and outline how robust inferential procedures may routinely be applied in practice in the biomediacal research. Numerical illustrations are given for the t-test, regression models, logistic regression, survival analysis and ROC curves, showing that robust methods are often more appropriate than standard procedures.

show abstract

Influence functions for proportional hazards regression

Cited by 87 publications

References 8 publications

Clinical, epidemiological, and spatial characteristics of Vibrio parahaemolyticus diarrhea and cholera in the urban slums of Kolkata, India

Clinical, epidemiological, and spatial characteristics of Vibrio parahaemolyticus diarrhea and cholera in the urban slums of Kolkata, India

Case-Weighted Measures of Influence for Proportional Hazards Regression

An overview of robust methods in medical research

Contact Info

Product

Resources

About