Diagnostics for the Cox model

Xue, Yishu; Schifano, Elizabeth D.

doi:10.29220/csam.2017.24.6.583

Cited by 36 publications

(32 citation statements)

References 83 publications

(96 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It assumes proportional hazards, an assumption that relative risks, such as male vs female, are constant over time. The proportional hazards assumption of the Cox regression model was assessed by chi‐square tests of Schoenfeld residuals …”

Section: Methodsmentioning

confidence: 99%

Treatment modality and outcomes in larynx cancer patients: A sex‐based evaluation

Kejner

2019

Head & Neck

View full text Add to dashboard Cite

Background To evaluate the differences in treatment modality and outcomes between male and female patients with laryngeal squamous cell carcinoma (SCC) in the United States. Methods Data were extracted from the Surveillance, Epidemiology, and End Results Database for patients with laryngeal SCC (2004‐2013). Overall survival (OS), disease specific survival (DSS), and multivariate analyses were conducted. Results Among women, supraglottic cancers were the most prevalent (61.2%); whereas the majority of men developed glottic cancers (64.2%; P < .001). Women with T4 disease were more likely to undergo primary radiation (56.8% vs 45.3%; P < .001) and less likely to undergo open surgery were than men (37.1% vs 48.2%; P < .001). Women had significantly better OS in glottic and supraglottic cancers, and comparable survival in subglottic cancer. Sex remained an independent prognostic factor for both DSS and OS. Conclusion Sex is an independent prognostic factor for DSS and OS in patients with laryngeal SCC.

show abstract

Section: Methodsmentioning

confidence: 99%

Treatment modality and outcomes in larynx cancer patients: A sex‐based evaluation

Kejner

2019

Head & Neck

View full text Add to dashboard Cite

show abstract

“…Common choices of

g (t)

include the Kaplan‐Meier (KM) transformation, which scales the horizontal axis by the left‐continuous version of the KM survival curve, the identity function, and the natural logarithm function. Formulation is rather general, as many tests fall within this framework for different choices of

g (t)

(see, eg, Xue and Schifano, ). Writing in matrix notation yields

λ_{i} (t) = λ_{0} (t) exp [X_{i} {(t)}^{⊤} {bold-italicβ + bold-italicG (t) bold-italicθ}], i = 1, …, n,

where

bold-italicG (t)

is a

p \times p

diagonal matrix with the j th diagonal element being

g_{j} (t)

, and

bold-italicθ = {(θ_{1}, …, θ_{p})}^{⊤}

.…”

Section: Cox Ph Modelmentioning

confidence: 99%

“…The online updating cumulative statistic

T_{k} (G)

was calculated to be 95.60. Due to the relatively high censoring rate, the KM transformation was chosen for calculation of the diagnostic statistics as it is more robust in such a scenario (eg, Xue and Schifano, ). Diagnosis with function

plot.cox.zph()

in the survival package revealed that all the parameters are likely to be time‐dependent; see Figure S5.…”

Section: Survival Analysis Of Seer Lymphoma Patientsmentioning

confidence: 99%

“…Common choices of g t ( ) include the Kaplan-Meier (KM) transformation, which scales the horizontal axis by the left-continuous version of the KM survival curve, the identity function, and the natural logarithm function. Formulation (4) is rather general, as many tests fall within this framework for different choices of g t ( ) (see, eg, Xue and Schifano, 2017). Writing (4) in matrix notation yields The test of Grambsch and Therneau (1994) is based on Schoenfeld residuals.…”

Section: Notation and Preliminariesmentioning

confidence: 99%

“…The Cox model (Cox, 1972) is the most commonly used tool in analyzing survival data. A crucial step in fitting the popular Cox model is to check the proportional hazards (PH) assumption (eg, Xue and Schifano, 2017). The standard approach, if new data become available along a stream, would be to pool all historical data together, fit a new Cox model, and use standard methods such as the test of Grambsch and Therneau (1994) to examine whether the PH assumption is appropriate.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An online updating approach for testing the proportional hazards assumption with streams of survival data

et al. 2019

Self Cite

View full text Add to dashboard Cite

The Cox model—which remains the first choice for analyzing time‐to‐event data, even for large data sets—relies on the proportional hazards (PH) assumption. When survival data arrive sequentially in chunks, a fast and minimally storage intensive approach to test the PH assumption is desirable. We propose an online updating approach that updates the standard test statistic as each new block of data becomes available and greatly lightens the computational burden. Under the null hypothesis of PH, the proposed statistic is shown to have the same asymptotic distribution as the standard version computed on an entire data stream with the data blocks pooled into one data set. In simulation studies, the test and its variant based on most recent data blocks maintain their sizes when the PH assumption holds and have substantial power to detect different violations of the PH assumption. We also show in simulation that our approach can be used successfully with “big data” that exceed a single computer's computational resources. The approach is illustrated with the survival analysis of patients with lymphoma cancer from the Surveillance, Epidemiology, and End Results Program. The proposed test promptly identified deviation from the PH assumption, which was not captured by the test based on the entire data.

show abstract

Efficacy of ground‐based trapping of coastal American black bears in Prince William Sound, Alaska

Westing

Skinner

Burcham

2022

Wildlife Society Bulletin

View full text Add to dashboard Cite

Ground-based trapping can be the most practical approach for capturing American black bears (Ursus americanus) in remote portions of Alaska. During the summers of 2016-2018 we captured 106 black bears in a coastal portion of southcentral Alaska, USA, using bucket cable traps and Aldrich snare sets, 2 common foot snaring techniques. Despite widespread use, little research has focused on factors contributing to success and differences in safety for both foot snaring techniques. We compared trap engagement and success and found that bears engaged with bucket cable traps at a 31% (CL = 3-50%) lower rate than with Aldrich snare sets but once engaged were equally likely to be successfully snared. Prebaiting sites increased capture success 362% (CL = 21-1663%) but it was unclear if success improved when sites were prebaited multiple times. Regardless of trap type, bear engagement with sets were 194% higher (CL = 94-345%) during mid-day hours (0800-1559) compared to crepuscular or nocturnal hours. We observed average injury scores of less than 11 points (out of 400 possible points) for both set types, consistent with the score given for minor lacerations.Both methods performed well within the recommended safety standards of the American Society of Mammalogists.

show abstract

Diagnostics for the Cox model

Cited by 36 publications

References 83 publications

Treatment modality and outcomes in larynx cancer patients: A sex‐based evaluation

Treatment modality and outcomes in larynx cancer patients: A sex‐based evaluation

An online updating approach for testing the proportional hazards assumption with streams of survival data

Efficacy of ground‐based trapping of coastal American black bears in Prince William Sound, Alaska

Contact Info

Product

Resources

About