-
[show abstract]
[hide abstract]
ABSTRACT: BackgroundThe focus of this study was the relative survival rates of breast cancer patients whose treatment was breast-conserving surgery
compared with that of mastectomy, adjusting for tumor size and nodal status because these factors may be intrinsically associated
with mastectomy being the treatment of choice. Patient age was also accounted for in the model. By adjusting for these factors,
we mitigate them as confounders of treatment choice in assessing effects on survival rates.
MethodsData were sourced from linked administrative data from the Western Australian Department of Health Record Linkage Unit. The
data consisted of linked records containing the diagnosis, subsequent hospital admission, and death records of about 3000
women diagnosed with cancer in Western Australia between 1 January 1995 and 31 December 1999. Cox proportional hazards regression
was used to investigate survival outcomes of breast-conserving surgery compared with that of mastectomy, adjusting for tumor
size, nodal status, and subject age.
ResultsThe hazard of death is reduced by a factor of about one half for subjects whose treatment was breast-conserving surgery over
treatment by mastectomy. Furthermore, the hazard of death increases substantially for subjects with nodal involvement over
subjects for whom there has been no identified spread to regional lymph nodes. Hazard of death increases as both age and tumor
size increase.
ConclusionsWestern Australian breast cancer patients treated with breast-conserving surgery have improved survival outcomes over those
treated with mastectomy, after allowing for tumor size, patient age, and lymph node involvement.
Annals of Surgical Oncology 04/2012; 14(1):157-164. · 4.17 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To investigate the extent to which standard errors can be underestimated in time-series studies of the association between particulate matter air pollution (PM) and mortality if model selection variation is not accounted for.
Actual-time series data from Cook County, Illinois, and Salt Lake County, Utah, for the period 1987 to 2000 were used to generate mortality time series. These series were used to examine the overconfidence resulting from ignoring variability introduced by the model selection process.
When variation associated with a model selection process is not accounted for, we found that the estimated standard errors for the effect of PM on mortality were substantially smaller than the true standard errors that necessarily incorporate model selection variability. Because of this, the true standard errors are approximately 70% larger than the reported standard errors. We also found that not accounting for model selection effects can result in the observed size of tests of no association between PM and mortality being up to about five times the nominal significance level.
Failing to account properly for the effect of model selection can reduce the accepted burden of proof for concluding a statistically significant association between PM and mortality.
Annals of epidemiology 10/2010; 20(10):772-8. · 2.95 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Single-case deletion regression diagnostics have been used widely to discover unusual data points, but such approaches can fail in the presence of multiple unusual data points and as a result of masking. We propose a new approach to the use of single-case deletion diagnostics that involves applying these diagnostics to delete-2 and delete-3 jackknife replicates of the data, and considering the percentage of times among these replicates that points are flagged as unusual as an indicator of their influence. By considering replicates that exclude certain collections of points, subtle masking effects can be uncovered.
Australian & New Zealand Journal of Statistics 02/2010; 52(1):45 - 60. · 0.44 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: We propose a bootstrap approach to gauging the size of regression influence measures. The bootstrap cut-offs generated are based on approximating the sampling distribution of the respective measures under resampling, work well for small samples, and allow for features such as asymmetric cut-offs. The bootstrap method uses Efron's jackknife-after-bootstrap idea to deal with the issue of an influential point contaminating the resamples from which cut-offs are calculated. The method is illustrated through both real-world examples and a simulation study, the results of which suggest that the bootstrap method provides a reliable alternative to traditional methods particularly in small to moderate samples.
Journal of Nonparametric Statistics 02/2010; 22(2):257-269. · 0.46 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Concerns have been raised about findings of associations between particulate matter (PM) air pollution and mortality that have been based on a single "best" model arising from a model selection procedure, because such a strategy may ignore model uncertainty inherently involved in searching through a set of candidate models to find the best model. Model averaging has been proposed as a method of allowing for model uncertainty in this context.
To propose an extension (double BOOT) to a previously described bootstrap model-averaging procedure (BOOT) for use in time series studies of the association between PM and mortality. We compared double BOOT and BOOT with Bayesian model averaging (BMA) and a standard method of model selection [standard Akaike's information criterion (AIC)].
Actual time series data from the United States are used to conduct a simulation study to compare and contrast the performance of double BOOT, BOOT, BMA, and standard AIC.
Double BOOT produced estimates of the effect of PM on mortality that have had smaller root mean squared error than did those produced by BOOT, BMA, and standard AIC. This performance boost resulted from estimates produced by double BOOT having smaller variance than those produced by BOOT and BMA.
Double BOOT is a viable alternative to BOOT and BMA for producing estimates of the mortality effect of PM.
Environmental Health Perspectives 01/2010; 118(1):131-6. · 7.04 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The authors propose a regression-based approach for obtaining multiday estimates of the adverse health effects of ambient particulate matter less than 10 microm in diameter (PM(10)) when daily PM(10) time-series data are unavailable. This situation is common in the United States, because most US cities take PM(10) measurements every 6 days. Current evidence suggests that adverse effects of PM(10) are not concentrated on a single day but rather are spread out over multiple days, so the unavailability of daily PM(10) data presents a problem for the estimation of these effects. The proposed model estimates weights that are used to construct a linear combination of single-lag PM(10) effect estimates obtained from the available PM(10) data. It is shown that this new approach provides estimates of the effect of PM(10) on mortality that have less bias and mean squared error than currently available methods. Application of this method to the US cities contained in the National Morbidity, Mortality, and Air Pollution Study database produces an estimated national average effect of PM(10) on nonaccidental mortality in persons over age 65 years, corresponding to a 0.32% increase per 10-microg/m(3) increment in PM(10). The estimated effects for cardiorespiratory mortality and other mortality are 0.34% and 0.22%, respectively.
American journal of epidemiology 07/2008; 167(12):1511-7. · 5.59 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: In time-series studies on the effect of particulate matter (PM) air pollution on an adverse health outcome, PM time-series data are often available from multiple monitoring stations. Published studies have combined the data from the multiple monitors using a simple or trimmed average. We investigate an alternative method of combining the data available from multiple PM-monitoring sites. This method uses time-series data to assign each PM monitor a weight. The weights are then used to combine the data from the multiple PM monitors into a single air pollution time series. The resulting model will identify important monitors for describing the relationship between PM and the adverse health outcome of interest. Subsequent investigations of why certain monitors are more informative than others may provide valuable information concerning the location of vulnerable subpopulations or locations where the meteorological and/or land-use conditions are better for assessing population exposure to PM. The new model is illustrated by applying it to actual data from Cook County, IL, USA and through a simulation study. Using the new model, for the Cook County data, it was found that two of the six monitors provided essentially as much information about the effect of PM on mortality as all six monitors combined. The simulation study suggests that the weights assigned to each monitor by the new model are appropriate, that is, that the model assigns the largest weight to the monitor most highly correlated with the underlying PM time series used to generate mortality.
Journal of Exposure Science and Environmental Epidemiology 02/2008; 18(1):88-94. · 2.93 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The majority of studies that have investigated the relationship between particulate matter (PM) air pollution and mortality have assumed a linear dose-response relationship and have used either a single-day's PM or a 2- or 3-day moving average of PM as the measure of PM exposure. Both of these modeling choices have come under scrutiny in the literature, the linear assumption because it does not allow for non-linearities in the dose-response relationship, and the use of the single- or multi-day moving average PM measure because it does not allow for differential PM-mortality effects spread over time. These two problems have been dealt with on a piecemeal basis with non-linear dose-response models used in some studies and distributed lag models (DLMs) used in others. In this paper, we propose a method for investigating the shape of the PM-mortality dose-response relationship that combines a non-linear dose-response model with a DLM. This combined model will be shown to produce satisfactory estimates of the PM-mortality dose-response relationship in situations where non-linear dose response models and DLMs alone do not; that is, the combined model did not systemically underestimate or overestimate the effect of PM on mortality. The combined model is applied to ten cities in the US and a pooled dose-response model formed. When fitted with a change-point value of 60 microg/m(3), the pooled model provides evidence for a positive association between PM and mortality. The combined model produced larger estimates for the effect of PM on mortality than when using a non-linear dose-response model or a DLM in isolation. For the combined model, the estimated percentage increase in mortality for PM concentrations of 25 and 75 microg/m(3) were 3.3% and 5.4%, respectively. In contrast, the corresponding values from a DLM used in isolation were 1.2% and 3.5%, respectively.
Environmental Research 07/2007; 104(2):193-200. · 3.40 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: In many cities of the United States, measurements of ambient particulate matter air pollution (PM) are available only every sixth day. Time-series studies conducted in these cities that investigate the relationship between mortality and PM are restricted to using a single day's PM as the measure of PM exposure, rather than using measurements taken over several consecutive days. Studies showed that using a single-day PM as the measure of PM exposure can result in estimates that have a negative bias, sometimes in the order of over half of the value being estimated. In this article two methods are introduced that can be used to obtain estimates that can in some situations reduce the bias to negligible proportions when only every-sixth-day PM concentrations are available. Using one of these methods, the national average PM mortality effect estimates obtained for total mortality and cardiovascular and respiratory mortality, respectively, correspond to 0.27% and 0.39% increases in mortality per 10-microg/m3 increment in PM. The corresponding effect estimates obtained using the single-day lag-1 PM concentration are 0.18% and 0.23%. The estimates obtained using the lag-1 PM concentration were the most widely reported results from the recent multicity National Morbidity, Mortality, and Air Pollution Study (NMMAPS) analyses. The more accurate estimates obtained from the methods introduced in this article will enable more accurate quantification of the increased incidence in mortality due to elevation in PM levels and the benefit of current or more stringent regulatory standards.
Journal of Toxicology and Environmental Health Part A 05/2007; 70(8):665-75. · 1.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Unplanned hospital readmissions after surgical treatment for breast cancer are an indicator of morbidity. We explore the relationship between the rate of unplanned hospital readmissions within 42 days of initial treatment and various factors, including tumor size and histology, lymph node involvement, type of surgical treatment, mastectomy, or breast-conserving surgery, and patient demographics.
Linked Western Australian cancer mortality and hospital morbidity data were used in the assessment of readmissions within a period of 42 days after initial surgical treatment for breast cancer. Planned admissions for adjuvant treatment such as chemotherapy or radiotherapy were deleted. Survival models for multiple events per subject were applied to analyze the data.
The analysis reveals that patients more likely to experience lower recurrence of short-term unplanned hospital readmissions include those with smaller tumors, private insurance, and who reside in metropolitan areas. The model also includes important two-way interaction terms involving tumor histology, area of residence, and surgical treatment, and between lymph node involvement and patient age.
This study suggests that the choice of breast-conserving surgery as a treatment for breast cancer does not invariably result in better postoperative morbidity, but rather, that other factors, including tumor size and patient demographics, play a critical role in the short term. These results differ from a previous study of longterm hospital readmissions-country of birth and method of payment were found to be associated with short-term hospital admission but not with longterm readmissions.
Journal of the American College of Surgeons 03/2007; 204(2):193-200. · 4.55 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The focus of this study was the relative survival rates of breast cancer patients whose treatment was breast-conserving surgery compared with that of mastectomy, adjusting for tumor size and nodal status because these factors may be intrinsically associated with mastectomy being the treatment of choice. Patient age was also accounted for in the model. By adjusting for these factors, we mitigate them as confounders of treatment choice in assessing effects on survival rates.
Data were sourced from linked administrative data from the Western Australian Department of Health Record Linkage Unit. The data consisted of linked records containing the diagnosis, subsequent hospital admission, and death records of about 3000 women diagnosed with cancer in Western Australia between 1 January 1995 and 31 December 1999. Cox proportional hazards regression was used to investigate survival outcomes of breast-conserving surgery compared with that of mastectomy, adjusting for tumor size, nodal status, and subject age.
The hazard of death is reduced by a factor of about one half for subjects whose treatment was breast-conserving surgery over treatment by mastectomy. Furthermore, the hazard of death increases substantially for subjects with nodal involvement over subjects for whom there has been no identified spread to regional lymph nodes. Hazard of death increases as both age and tumor size increase.
Western Australian breast cancer patients treated with breast-conserving surgery have improved survival outcomes over those treated with mastectomy, after allowing for tumor size, patient age, and lymph node involvement.
Annals of Surgical Oncology 02/2007; 14(1):157-64. · 4.17 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: BACKGROUND AND OBJECTIVES: Unplanned hospital readmissions following surgical treatment for breast cancer are an indicator of morbidity. We explore the relationship between the risk of unplanned hospital readmissions and various factors, including tumor size and histology, lymph node involvement, the type of surgical treatment, mastectomy or breast-conserving surgery (BCS), and patient demographics. METHODS: Linked Western Australian cancer mortality and hospital morbidity data were used in the assessment of readmissions following initial surgical treatment for breast cancer. Planned admissions for adjuvant treatment such as chemotherapy or radiotherapy were deleted. Survival models for multiple events per subject were applied to analyze the data. RESULTS: Hazard of unplanned readmission rises by a factor of 1.005 for each mm in tumor size, is reduced by about 40% for metropolitan residents over rural-based patients, and by 4% for patients whose initial surgical treatment was BCS over mastectomy patients. Area of residence interacts with other factors, including patient age, lymph node involvement, and tumor histology. CONCLUSIONS: While use of BCS appears associated with lower long-term rates of unplanned hospital readmissions than those following mastectomy, the roles of other factors remain important. Patients living in metropolitan areas have significantly lower rates of readmission than rural/remote counterparts. J. Surg. Oncol. (c) 2007 Wiley-Liss, Inc.
Journal of Surgical Oncology 02/2007; · 2.10 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Many investigations of the adverse health effects of multiple air pollutants analyze the time series involved by simultaneously entering the multiple pollutants into a Poisson log-linear model. This method can yield unstable parameter estimates when the pollutants involved suffer high intercorrelation; therefore, traditional approaches to dealing with multicollinearity, such as principal component analysis (PCA), have been promoted in this context.
A characteristic of PCA is that its construction does not consider the relationship between the covariates and the adverse health outcomes. A refined version of PCA, supervised principal components analysis (SPCA), is proposed that specifically addresses this issue.
Models controlling for longterm trends and weather effects were used in conjunction with each SPCA and PCA to estimate the association between multiple air pollutants and mortality for U.S. cities. The methods were compared further via a simulation study.
Simulation studies demonstrated that SPCA, unlike PCA, was successful in identifying the correct subset of multiple pollutants associated with mortality. Because of this property, SPCA and PCA returned different estimates for the relationship between air pollution and mortality.
Although a number of methods for assessing the effects of multiple pollutants have been proposed, such methods can falter in the presence of high correlation among pollutants. Both PCA and SPCA address this issue. By allowing the exclusion of pollutants that are not associated with the adverse health outcomes from the mixture of pollutants selected, SPCA offers a critical improvement over PCA.
Environmental Health Perspectives 01/2007; 114(12):1877-82. · 7.04 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The shape of the dose-response relation between particulate matter air pollution and mortality is crucial for public health assessment, and departures of this relation from linearity could have important regulatory consequences. A number of investigators have studied the shape of the particulate matter-mortality dose-response relation and concluded that the relation could be adequately described by a linear model. Some of these researchers examined the hypothesis of linearity by comparing Akaike's Information Criterion (AIC) values obtained under linear, piecewise linear, and spline alternative models. However, at the current time, the efficacy of the AIC in this context has not been assessed. The authors investigated AIC as a means of comparing competing dose-response models, using data from Cook County, Illinois, for the period 1987-2000. They found that if nonlinearities exist, the AIC is not always successful in detecting them. In a number of the scenarios considered, AIC was equivocal, picking the correct simulated dose-response model about half of the time. These findings suggest that further research into the shape of the dose-response relation using alternative model selection criteria may be warranted.
American Journal of Epidemiology 01/2007; 164(12):1242-50. · 5.22 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This paper studies properties of the portmanteau statistic proposed by Box and Pierce 11.
Box , G. and Pierce , D. 1970 . Distribution of residual autocorrelations in autoregressive-integrated moving average time series models . Journal of the American Statistical Association , 65 : 1509 – 1526 . [CrossRef]View all references and its modification of Ljung and Box 22.
Ljung , G. and Box , G. 1978 . On a measure of lack of fit in time series models . Biometrika , 65 : 297 – 303 . [CrossRef], [Web of Science ®]View all references. We show that these portmanteau statistics are feasible analogs to optimal tests for the class of statistics which are linear combinations of consistent estimates of serial correlations. We find, however, that for sample sizes commonly encountered in practice, the efficiency loss in power of portmanteau statistics relative to optimal tests can be substantial, although their size properties are broadly comparable. Our results indicate that tests based on some other non-optimal weighting schemes, including tests with optimal weights constructed from moderately misspecified alternatives, deliver tests with better power than the Box–Pierce or Ljung–Box statistics.
Journal of Statistical Computation and Simulation 07/2006; 76(7):593-604. · 0.50 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The consensus from time series studies that have investigated the mortality effects of particulate matter air pollution (PM) is that increases in PM are associated with increases in daily mortality. However, recently concerns have been raised that the observed positive association between PM and mortality may be an artefact of model selection due to multiple hypothesis testing. This problem arises when a number of models are investigated, but only the "best" model is reported and all subsequent inference is based on this model, ignoring the model selection process. In this paper, we introduce the use of the bootstrap as a means of addressing the problems of model selection in PM mortality time series studies. Using the bootstrap to perform inference about the effect of PM on mortality is a process based on a set of models rather than on a single model. It is shown that using the bootstrap to overcome the problems of model selection is competitive with the existing methodology of Bayesian model averaging.
Journal of Exposure Science and Environmental Epidemiology 06/2006; 16(3):242-50. · 2.93 Impact Factor
-
Steven Roberts
[show abstract]
[hide abstract]
ABSTRACT: Because the U.S. Environmental Protection Agency regulates air pollutants independently, the majority of time-series studies on air pollution and mortality have focused on estimating the adverse health effects of a single pollutant. However, due to the sometimes high correlation between air pollutants, the results from studies that focus on a single air pollutant can be difficult to interpret. In addition, the high correlation between air pollutants can produce problems of interpretation for the standard method of investigating the adverse health effects due to multiple air pollutants. The standard method involves simultaneously including the multiple air pollutants in a single statistical model. Because of this, the development of new models to concurrently estimate the adverse health effects of multiple air pollutants has recently been identified as an important area of future research. In this article, a new model for disentangling the joint effects of multiple air pollutants in air pollution mortality time-series studies is introduced. This new model uses the time-series data to assign each air pollutant a weight that indicates the pollutant's contribution to the air pollution mixture that affects mortality and to estimate the effect of this air pollution mixture on mortality. This model offers an improvement in statistical estimation precision over the standard method. It also avoids problems of interpretation that can occur if the standard method is used. This new model is then illustrated by applying it to time-series data from two U.S. counties.
Journal of Toxicology and Environmental Health Part A 04/2006; 69(6):417-35. · 1.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: A critical choice facing breast cancer patients is which surgical treatment--mastectomy or breast conserving surgery (BCS)--is most appropriate. Several studies have investigated factors that impact the type of surgery chosen, identifying features such as place of residence, age at diagnosis, tumor size, socio-economic and racial/ethnic elements as relevant. Such assessment of "propensity" is important in understanding issues such as a reported under-utilisation of BCS among women for whom such treatment was not contraindicated. Using Western Australian (WA) data, we further examine the factors associated with the type of surgical treatment for breast cancer using a classification tree approach. This approach deals naturally with complicated interactions between factors, and so allows flexible and interpretable models for treatment choice to be built that add to the current understanding of this complex decision process.
Data was extracted from the WA Cancer Registry on women diagnosed with breast cancer in WA from 1990 to 2000. Subjects' treatment preferences were predicted from covariates using both classification trees and logistic regression.
Tumor size was the primary determinant of patient choice, subjects with tumors smaller than 20 mm in diameter preferring BCS. For subjects with tumors greater than 20 mm in diameter factors such as patient age, nodal status, and tumor histology become relevant as predictors of patient choice.
Classification trees perform as well as logistic regression for predicting patient choice, but are much easier to interpret for clinical use. The selected tree can inform clinicians' advice to patients.
BMC Cancer 02/2006; 6:98. · 3.01 Impact Factor
-
Steven Roberts
[show abstract]
[hide abstract]
ABSTRACT: In many cities of the United States, measurements of ambient particulate matter air pollution (PM) are available only once every 6 days. Time-series studies conducted in these cities that investigate the relationship between mortality and PM are restricted to using a single day's PM as the measure of PM exposure. This is undesirable because current evidence suggests that the effects of PM on mortality are spread over multiple days. And studies have shown that using a single day's PM as the measure of PM exposure can result in estimates that have a large negative bias. In this article, I introduce a new model for estimating the mortality effects of PM when only every-sixth-day PM data are available. This new model uses information available in the daily mortality time series to infer otherwise lost information about the effect of PM on mortality over a period of more than a single day. This new model typically offers an increase in both statistical estimation precision and accuracy compared with existing models.
Environmental Health Perspectives 10/2005; 113(9):1148-52. · 7.04 Impact Factor
-
Steven Roberts
[show abstract]
[hide abstract]
ABSTRACT: In particulate air pollution mortality time series studies, the particulate air pollution exposure measure used is typically the current day's or the previous day's air pollution concentration or a multi-day moving average air pollution concentration. Distributed lag models (DLMs) that allow for differential air pollution effects that are spread over multiple days are seen as an improvement over using a single- or multi-day moving average air pollution exposure measure. However, at the current time, the statistical properties of DLMs as a measure of air pollution exposure have not been investigated. In this paper, a simulation study is used to investigate the performance of DLMs as a measure of air pollution exposure in comparison with single- and multi-day moving average air pollution exposure measures under various forms for the true effect of air pollution on mortality. The simulation study shows that DLMs offer a more robust measure of the effect of air pollution on mortality and avoid the potential for a large negative bias compared with single- or multi-day moving average air pollution exposure measures. This is important information. In many U.S. cities, particulate air pollution concentrations are observed only once every six days, meaning it is often only possible to use single-day particulate air pollution exposure measures. The results from this paper will help quantify the magnitude of the negative bias that can result from using single-day exposure measures. The implications of this work for future air pollution mortality time series studies are discussed. The data used in this paper are concurrent daily time series of mortality, weather, and particulate air pollution from Cook County, IL, for the period 1987-1994.
Journal of the Air & Waste Management Association (1995) 04/2005; 55(3):273-82. · 1.52 Impact Factor
