<i>partDSA</i>: deletion/substitution/addition algorithm for partitioning the covariate space in prediction

Molinaro, Annette M.; Lostritto, Karen; Laan, Mark J. van der

doi:10.1093/bioinformatics/btq142

Cited by 40 publications

(26 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Analysis by the partitioning deletion/substitution/addition algorithm (partDSA) was used to determine the cutoff points for the timing of chemoradiation initiation at which there was a difference in OS and PFS. 16,17 The log-rank test was used for comparison of survival between groups. The Cox proportional hazards model was used to assess the effect of timing of concurrent chemoradiation (in days) on outcomes, while adjusting for variables: treatment protocol, age, Karnofsky Performance Score (KPS), and extent of resection.…”

Section: Methodsmentioning

confidence: 99%

The Effect of Timing of Concurrent Chemoradiation in Patients With Newly Diagnosed Glioblastoma

et al. 2015

Self Cite

View full text Add to dashboard Cite

BACKGROUND The effect of timing of initiation of concurrent radiation and chemotherapy after surgery on outcome of patients with glioblastoma (GBM) remains unclear. OBJECTIVE To further explore this issue, we analyzed 4 clinical trials for patients newly diagnosed with GBM receiving concurrent and adjuvant temozolomide. METHODS The cohort study included 198 adult patients with newly diagnosed supratentorial GBM who were enrolled from 2004 to 2010 in 4 clinical trials consisting of radiation plus temozolomide and an experimental agent. The interval to initiation of therapy was determined from the time of surgical resection. The partitioning deletion/substitution/addition algorithm was used to determine the cutoff points for timing of chemoradiation at which there was a significant difference in overall survival (OS) and progression-free survival (PFS). RESULTS The median wait time between surgery and initiation of concurrent chemoradiation was 29.5 days (range, 7–56 days). A short delay in chemoradiation administration (at 30–34 days) was predictive of prolonged OS (hazard ratio [HR]: 0.63, P = .03) and prolonged PFS (HR: 0.68, P = .06) compared with early initiation of concurrent chemoradiation (<30 days), after adjusting for protocol and baseline prognostic variables including extent of resection by multivariate analysis. A longer delay to chemoradiation beyond 34 days was not associated with improved OS or PFS compared with early initiation (HR: 0.94, P = .77 and HR: 0.91, P = .63, respectively). CONCLUSION A short delay in the start of concurrent chemoradiation is beyond the classic paradigm of 4 weeks post-resection and may be associated with prolonged OS and PFS.

show abstract

Section: Methodsmentioning

confidence: 99%

The Effect of Timing of Concurrent Chemoradiation in Patients With Newly Diagnosed Glioblastoma

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…The partitioning deletion/substitution/addition algorithm was used to determine the cutoff points for timing of chemoradiation at which there was a significant difference in OS and PFS. 23 The partitioning deletion/substitution/addition algorithm was developed for use in cancer survival analyses that use multiple predictive variables and their interactions. The algorithm chooses the best among all “or” scenarios of variables while allowing for both continuous and categorical variables in the model.…”

Section: Published Literaturementioning

confidence: 99%

Impact of Timing of Concurrent Chemoradiation for Newly Diagnosed Glioblastoma

et al. 2015

Self Cite

View full text Add to dashboard Cite

“…In particular, because partDSA iterates among three possible moves and performs an exhaustive search of the covariate space, it will inevitably require a significantly higher running time than CART. The R package for partDSA allows for the cross‐validation folds to be run in parallel, making runnings time feasible in many applications (Molinaro et al, 2010), especially given that most clinical data sets remain relatively modest in size. In addition to extensions mentioned previously, future work on partDSA includes further work on methods for variable selection and associated variable importance measures as well as methods for increasing computational efficiency.…”

Section: Discussionmentioning

confidence: 99%

“…Molinaro, Lostritto, and van der Laan (2010) recently developed partDSA , a new loss‐based recursive partitioning method. Like CART, partDSA divides the covariate space into mutually exclusive and disjoint regions on the basis of a chosen loss function.…”

Section: Introductionmentioning

confidence: 99%

“…For example, with a continuous outcome and an L 2 (i.e., squared-error) loss function, the predicted value for each terminal node is the mean outcome for all observations falling into that node. Molinaro, Lostritto, and van der Laan (2010) recently developed partDSA, a new loss-based recursive partitioning method. Like CART, partDSA divides the covariate space into mutually exclusive and disjoint regions on the basis of a chosen loss function.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Partitioning Deletion/Substitution/Addition Algorithm for Creating Survival Risk Groups

Lostritto

Strawderman

2012

Biometrics

Self Cite

View full text Add to dashboard Cite

Accurately assessing a patient's risk of a given event is essential in making informed treatment decisions. One approach is to stratify patients into two or more distinct risk groups with respect to a specific outcome using both clinical and demographic variables. Outcomes may be categorical or continuous in nature; important examples in cancer studies might include level of toxicity or time to recurrence. Recursive partitioning methods are ideal for building such risk groups. Two such methods are Classification and Regression Trees (CART) and a more recent competitor known as the partitioning Deletion/Substitution/Addition (partDSA) algorithm, both which also utilize loss functions (e.g. squared error for a continuous outcome) as the basis for building, selecting and assessing predictors but differ in the manner by which regression trees are constructed.Recently, we have shown that partDSA often outperforms CART in so-called "full data" (e.g., uncensored) settings. However, when confronted with censored outcome data, the loss functions used by both procedures must be modified. There have been several attempts to adapt CART for right-censored data. This article describes two such extensions for partDSA that make use of observed data (i.e. possibly censored) loss functions. These observed data loss functions, constructed using inverse probability of censoring weights, are consistent estimates of their uncensored counterparts provided that the corresponding censoring model is correctly specified. The relative performance of these new methods is evaluated via simulation studies and illustrated through an analysis of clinical trial data on brain cancer patients. The implementation of partDSA for uncensored and right censored outcomes is publicly available in the R package, partDSA.

show abstract

partDSA: deletion/substitution/addition algorithm for partitioning the covariate space in prediction

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 40 publications

References 6 publications

The Effect of Timing of Concurrent Chemoradiation in Patients With Newly Diagnosed Glioblastoma

The Effect of Timing of Concurrent Chemoradiation in Patients With Newly Diagnosed Glioblastoma

Impact of Timing of Concurrent Chemoradiation for Newly Diagnosed Glioblastoma

A Partitioning Deletion/Substitution/Addition Algorithm for Creating Survival Risk Groups

Contact Info

Product

Resources

About