Nonparametric Bootstrap Techniques Research Articles

Predictability of carbon emissions from biomass burning in Indonesia from 1997 to 2006

Drought and sea surface temperature were examined as the causes of severe biomass burning C emissions in Indonesia for 1997–2006, obtained from the Global Fire Emissions Database. Eighteen predictor variables were considered under log linear and piecewise regression models. The predictor variables considered were precipitation totals of up to 6 months, output from two soil moisture models, and sea surface temperature (SST) indicators reflecting El Niño and Indian Ocean Dipole strength. Nonparametric bootstrap techniques were used to estimate confidence intervals for predictability and thresholds below which severe C emissions are likely. Across equatorial Southeast Asia, the best predictor was 3‐month total precipitation, which explained 79% of variance in C emissions. When considered individually, and with the incorporation of satellite precipitation estimates, predictability for southern Sumatra and southern Kalimantan improved to 97% and 92%, respectively, using 4‐month total precipitation. There is a high risk of severe burning when 4‐month precipitation falls below thresholds of 350 mm in southern Sumatra and 650 mm in southern Kalimantan and when 6‐month precipitation falls below 900 mm in Papua. In general, simple precipitation totals outperformed more complicated soil moisture models and SST‐based indices. Physically, seasonal precipitation controls fire emissions through its regulation of groundwater level and, hence, the amount of peat available for drying. Seasonal precipitation, in turn, is strongly influenced by SST patterns in the tropical Pacific and Indian oceans. The most severe drought and fire events appear equally influenced by Indian Ocean Dipole events and El Niño events.

Determination of Transmission Reliability Margin Using Parametric Bootstrap Technique

The determination of available transfer capability (ATC) must accommodate a reasonable range of transmission reliability margin (TRM) so that the transmission network is secure from uncertainty of transfer capability that may occur during a power transfer. This paper presents a computationally accurate method in determining the TRM with large amount of transfer capability uncertainty using the parametric bootstrap technique. The parametric bootstrap technique is used to randomly generate a bootstrap sample of ATCs with large uncertainty selected at a certain percentage of bootstrap confidence interval. The bootstrap sample of ATCs in a day is used to determine the TRM at every time interval. The TRM value at a particular time interval is selected based on a certain percentage of normal cumulative distribution function (CDF). Then, a new value of ATC at the current time interval is calculated by considering the TRM at the same time interval. The effectiveness of the proposed TRM method in providing new ATC value is validated on the Malaysian power system. The results show that the proposed method provides accurate estimation of TRM in which it is relatively similar to the TRM result obtained by the standard deviation of uncertainty which is incorporated in the Monte Carlo simulation technique. Further comparisons in terms of accuracy and total computational time in estimating the TRM is made between the parametric and nonparametric bootstrap techniques.

Cost-effectiveness analysis of elective endovascular repair compared with open surgical repair of abdominal aortic aneurysms for patients at a high surgical risk: A 1-year patient-level analysis conducted in Ontario, Canada

Abdominal aortic aneurysm (AAA) is a prevalent health condition affecting up to 14% of men and 6% of women. The objective of this study was to estimate the cost-effectiveness and cost-utility of elective endovascular aneurysm repair (EVAR) compared with open surgical repair (OSR) in patients at a high risk of surgical complications. Patient-level cost and outcome data from a 1-year prospective observational study conducted at London Health Sciences Centre, London, Ontario, Canada, was used to determine the incremental cost per life-year gained and the incremental cost per quality-adjusted life year (QALY) gained of EVAR compared with OSR in patients with an AAA >5.5 cm and a high risk of surgical complications. The analysis was taken from a societal perspective and the time horizon was 1 year. To measure sampling uncertainty on costs and effects, nonparametric bootstrap techniques were applied. Uncertainty results were expressed using cost-effectiveness acceptability curves. Extrapolations of the 1-year results to a 5-year time horizon were conducted in sensitivity analyses. Between August 11, 2003, and April 3, 2005, 192 patients at a high risk of surgical complications were enrolled: 140 received EVAR and 52 OSR. Point estimates during a 1-year period showed that EVAR dominated OSR for high-risk patients in terms of incremental cost per life-year gained and incremental cost per QALYs. However, bootstrap estimates for the two cost-effectiveness measures indicated there was a great deal of uncertainty regarding the costs and the QALYs and less uncertainty regarding life-years gained. If society was willing to pay $50,000 per life-year gained or per QALY gained, the probability of EVAR being cost-effective was found to be 0.76 and 0.55, respectively. Five-year extrapolations indicated that EVAR was cost-effective compared with OSR. According to this 1-year observational study, EVAR may be a cost-effective strategy compared with OSR for high-risk patients. Longer-term data are needed to decrease the uncertainty associated with the results.

Cost analysis of amlodipine versus enalapril in patients with coronary artery disease and normal blood pressure

To analyse 2-year hospitalization and cost data collected during a prospective, double-blind, randomized, controlled trial comparing amlodipine, enalapril and placebo in normotensive patients with coronary artery disease (CAD). All patients who were enrolled in the CAMELOT study were included in this economic substudy. Patients with CAD and normal blood pressure were randomized to amlodipine, enalapril or placebo, and followed up for 24 months (between 1999 and 2004). Data on hospitalizations and medication use were obtained from the clinical trial. Costs were assigned from secondary sources. Total costs ($US, year 2004 values) were estimated as the sum of costs associated with cardiovascular hospitalizations, study medications and concomitant cardiovascular medications. Costs and resource use were analysed by treatment arm overall and for selected patient subgroups. Cost differences were evaluated using nonparametric bootstrap techniques. Of 1991 patients enrolled, 663 were treated with amlodipine, 673 were treated with enalapril and 655 were treated with placebo. Significantly fewer patients were hospitalized for cardiovascular reasons in the amlodipine group (16.4%) than in the placebo group (22.7%; p < 0.01), but not compared with the enalapril group (20.1%; p = 0.09). The amlodipine group also had numerically fewer days in hospital per patient (1.1) than the enalapril (1.3) and placebo (1.5) groups. Mean 2-year per-patient costs in the amlodipine group were estimated to be $US 609 and $US 717 lower than for the placebo and enalapril groups, respectively. These results suggest that use of amlodipine may reduce costs of care among CAD patients with normal blood pressure.

Cost-utility of an 8-month aquatic training for women with fibromyalgia: a randomized controlled trial

IntroductionPhysical therapy in warm water has been effective and highly recommended for persons with fibromyalgia, but its efficiency remains largely unknown. Should patients or health care managers invest in this therapy? The aim of the current study was to assess the cost-utility of adding an aquatic exercise programme to the usual care of women with fibromyalgia.MethodsCosts to the health care system and to society were considered in this study that included 33 participants, randomly assigned to the experimental group (n = 17) or a control group (n = 16). The intervention in the experimental group consisted of a 1-h, supervised, water-based exercise sessions, three times per week for 8 months. The main outcome measures were the health care costs and the number of quality-adjusted life-years (QALYs) using the time trade-off elicitation technique from the EuroQol EQ-5D instrument. Sensitivity analyses were performed for variations in staff salary, number of women attending sessions and time spent going to the pool. The cost effectiveness acceptability curves were created using a non-parametric bootstrap technique.ResultsThe mean incremental treatment costs exceeded those for usual care per patient by € 517 for health care costs and € 1,032 for societal costs. The mean incremental QALY associated with the intervention was 0.131 (95% CI: 0.011 to 0.290). Each QALY gained in association with the exercise programme cost an additional € 3,947/QALY (95% CI: 1,782 to 47,000) for a health care perspective and € 7,878/QALY (3,559 to 93,818) from a societal perspective. The curves showed a 95% probability that the addition of the water-based programme is a cost-effective strategy if the ceiling of inversion is € 14,200/QALY from a health care perspective and € 28,300/QALY from a societal perspective.ConclusionThe addition of an aquatic exercise programme to the usual care regime for fibromyalgia in women is cost effective in terms of both health care costs and societal costs. However, the characteristics of facilities (distance from the patients' homes and number of patients that can be accommodated per session) are major determinants to consider before investing in such a programme.Trial registrationCurrent controlled trials ISRCTN53367487.

Cost-effectiveness of drug-eluting stents in patients at high or low risk of major cardiac events in the Basel Stent KostenEffektivitäts Trial (BASKET): an 18-month analysis

Our aim was to determine whether drug-eluting stents are good value for money in long-term, everyday practice. We did an 18-month cost-effectiveness analysis of the Basel Stent KostenEffektivitäts Trial (BASKET), which randomised 826 patients 2:1 to drug-eluting stents (n=545) or to bare-metal stents (281). We used non-parametric bootstrap techniques to determine incremental cost-effectiveness ratios (ICERs) of drug-eluting versus bare-metal stents, to compare low-risk (> or =3.0 mm stents in native vessels; n=558, 68%) and high-risk patients (<3.0 mm stents/bypass graft stenting; n=268, 32%), and to do sensitivity analyses by altering costs and event rates in the whole study sample and in predefined subgroups. Quality-adjusted life-years (QALYs) were assessed by EQ-5D questionnaire (available in 703/826 patients). Overall costs were higher for patients with drug-eluting stents than in those with bare-metal stents (11,808 euros [SD 400] per patient with drug-eluting stents and 10,450 euros [592] per patient with bare-metal stents, mean difference 1358 euros [717], p<0.0001), due to higher stent costs. We calculated an ICER of 64,732 euros to prevent one major adverse cardiac event, and of 40,467 euros per QALY gained. Stent costs, number of events, and QALYs affected ICERs most, but unrealistic alterations would have been required to achieve acceptable cost-effectiveness. In low-risk patients, the probability of drug-eluting stents achieving an arbitrary ICER of 10,000 euros or less to prevent one major adverse cardiac event was 0.016; by contrast, it was 0.874 in high-risk patients. If used in all patients, drug-eluting stents are not good value for money, even if prices were substantially reduced. Drug-eluting stents are cost effective in patients needing small vessel or bypass graft stenting, but not in those who require large native vessel stenting.

Cost effectiveness of brace, physiotherapy, or both for treatment of tennis elbow

Background: The annual incidence of tennis elbow in the general population is high (1–3%). Tennis elbow often leads to limitation of activities of daily living and work absenteeism. Physiotherapy and...

The multiscale character of evoked cortical activity

Both the architecture and the dynamics of the brain have characteristic features at different spatial scales. However, the existence, nature and function of dynamical interdependencies between such scales have not been investigated. We studied the multiscale properties of functional magnetic resonance imaging (fMRI) data acquired while human subjects viewed a visual image. Traditional “region of interest” analysis of this data set revealed evoked activity in primary and extrastriate visual cortex. Wavelet transform in the spatial domain provides a multiscale representation of this evoked brain activity. Studying the correlation structure of this representation revealed strong and novel interdependencies in these data within and between different spatial scales. We found that such correlations are stronger than those evident in the original data and comparable in magnitude to those obtained after Gaussian smoothing. However, analysis of the data in the wavelet domain revealed additional structure such as positive correlations, strong anti-correlations and phase-lagged interdependencies. Statistical significance of these effects was inferred through nonparametric bootstrap techniques. We conclude that the spatial analysis of functional neuroimaging data in the wavelet domain provides novel information which may reflect complex spatiotemporal neuronal activity and information encoding. It also affords a quantitative means of testing hierarchical and multiscale models of cortical activity.

Impact of Changing the Statistical Methodology on Hospital and Surgeon Ranking

Risk adjustment is central to the generation of health outcome report cards. It is unclear, however, whether risk adjustment should be based on standard logistic regression, fixed-effects or random-effects modeling. The objective of this study was to determine how robust the New York State (NYS) Coronary Artery Bypass Graft (CABG) Surgery Report Card is to changes in the underlying statistical methodology. Retrospective cohort study based on data from the NYS Cardiac Surgery Reporting System on all patient undergoing isolated CABG surgery in NYS and who were discharged between 1997 and 1999 (51,750 patients). Using the same risk factors as in the NYS models, fixed-effects and random-effects models were fitted to the NYS data. Quality outliers were identified using 1) the ratio of observed-to-expected mortality rates (O/E ratio) and confidence intervals (CIs) calculated using both parametric (Poisson distribution) and nonparametric (bootstrapping) techniques; and 2) shrinkage estimators. At the surgeon level, the standard logistic regression model, the fixed-effects model, and the fixed-effects component of the random-effects model demonstrated near-perfect agreement on the identity of quality outliers using a quality indicator based on the O/E ratio and the Poisson distribution. Shrinkage estimators identified the fewest outliers, whereas the O/E ratios with bootstrap CI identified the greatest number of outliers. The results were similar for hospitals, except that the fixed-effects model identified more outliers than either the NYS model or the fixed-effects component of the random-effects model. Shrinkage estimators based on random-effects models are slightly more conservative in identifying quality outliers compared with the traditional approach based on fixed-effects modeling and standard regression. Explicitly modeling surgeon provider effect (fixed-effects and random-effects models) did not significantly alter the distribution of quality outliers when compared with standard logistic regression (which does not model provider effect). Compared with the standard parametric approach, the use of a bootstrap approach to construct 95% confidence interval around the O/E ratio resulted in more providers being identified as quality outliers.

Testing the Statistical Significance of Linear Programming Estimators

Linear programming–based estimation procedures are used in a variety of arenas. Two notable areas are multiattribute utility models (LINMAP) and production frontiers (data envelopment analysis (DEA)). Both LINMAP and DEA have theoretical and managerial advantages. For example, LINMAP treats ordinal-scaled preference data as such in uncovering individual-level attribute weights, while regression treats these preferences as interval scaled. DEA produces easy-to-understand efficiency measures, which allow for improved productivity benchmarking. However, acceptance of these techniques is hindered by the lack of statistical significance tests for their parameter estimates. In this paper, we propose and evaluate such parameter significance tests. Two types of tests are forwarded. The first examines whether a model’s fit is significantly reduced when an explanatory variable is deleted. The second is based on generating a standard deviation or distribution for the parameter estimate using nonparametric jackknife or bootstrap techniques. We demonstrate through simulations that both types of tests reliably identify both significant and insignificant parameters. The availability of these tests, especially the relatively simple and easy-to-use tests of the first type, should enhance the utilization of linear programming–based estimation.

Molecular testing of multiple HIV-1 transmissions in a criminal case

To test the a priori hypothesis of HIV-1 transmission from one suspect to six recipients in a criminal case. Partial pol and/or env sequences were obtained for at least two samples of the suspect and the victims. Appropriate local controls were sampled based on epidemiological and subtype criteria. Phylogenetic testing was performed using different reconstruction methods. Phylogenetic analyses consistently inferred a monophyletic cluster for the suspect and victim samples in both genome regions. This was highly supported by parametric and non-parametric bootstrapping techniques. Moreover, the controls most closely related to the suspect-victim cluster had a similar geographical origin to the suspect. Taking into account the limitations on the conclusions that can be drawn from molecular investigations we could infer that our molecular data is consistent with a scenario of multiple HIV transmission between suspect and victims.

Kosteneffizienzvergleich von Physiotherapie, manueller Therapie und hausärztlicher Therapie in der Behandlung von Nackenschmerzen

Objective: To evaluate the cost effectiveness of physiotherapy, manual therapy, and care by a general practitioner for patients with neck pain. Design: Economic evaluation alongside a randomized controlled trial. Setting: Primary care. Participants: A total of 183 patients with neck pain for at least two weeks recruited by 42 general practitioners and randomly allocated to manual therapy (n=60, spinal mobilization), physiotherapy (n=59, mainly exercise), or general practitioner care (n=64, counseling, education, and drugs). Main outcome measures: Clinical outcomes were perceived: recovery, intensity of pain, functional disability, and quality of life. Direct and indirect costs were measured by means of cost diaries that were kept by patients for one year. Differences in mean costs between groups, cost effectiveness, and cost utility ratios were evaluated by applying non-parametric bootstrapping techniques. Results: The manual therapy group showed a faster improvement than the physiotherapy group and the general practitioner care group up to 26 weeks, but differences were negligible by follow-up at 52 weeks. The total costs of manual therapy (447 EUR) were around one third of the costs of physiotherapy (1297 EUR) and general practitioner care (1379 EUR). These differences were significant:p<0.01 for manual therapy vs. physiotherapy and manual therapy vs. general practitioner care, and p=0.55 for general practitioner care vs. physiotherapy. The cost effectiveness ratios and the cost utility ratios showed that manual therapy was less costly and more effective than physiotherapy or general practitioner care. Conclusions: Manual therapy (spinal mobilization) is more effective and less costly for treating neck pain than physiotherapy or care by a general practitioner.

USE OF THE BOOTSTRAP FOR ERROR PROPAGATION IN ESTIMATING ZOOPLANKTON PRODUCTION

Error propagation is the calculation of statistical error in a quantity that comprises multiple components each with associated error. Such quantities constitute the end point of many ecological studies, but composite errors are rarely incorporated so that uncertainty in the final estimate is either unknown or underestimated. In this study we present the use of both parametric (observations are resampled from standard probability distributions) and nonparametric (raw observations are resampled) bootstrap techniques to propagate errors through the many steps involved in the egg-ratio estimation of seasonal production for the freshwater zooplankter Bythotrephes longimanus. We first compute parametric and nonparametric bootstrap estimates of the standard deviation in seasonal production of B. longimanus, showing that it ranges from 21% to 27% of observed seasonal production. We demonstrate that our bootstrapping procedures are robust by developing a theoretical model and showing that the true variance of a parameter is included in 96% of the simulated 95% confidence intervals. We also show that the choice of probability distribution in the parametric bootstrap can change the standard deviation of seasonal production by up to 90%. We argue that ecologists should use such error propagation techniques more routinely than is currently the case.

Uncertainties in sampling procedures for age composition of hake and sardine in Iberian Atlantic waters

Estimates of the age composition of hake (demersal) and sardine (pelagic) in Iberian Atlantic waters in 1999 are analysed, paying specific attention to sampling variability. Two methods are applied, one based on an analytical approach and the other on simulation using a non-parametric bootstrap technique. The methods were used to estimate the catch at age by numbers per species for 1999, and the associated variability. They provide a useful means of analysing the precision of the estimates, of assessing data quality, and of optimising different sampling schemes. The estimates are precise for the reference ages used in assessment (CV < 13% for both species). The bootstrap method is considered better owing to its non-parametric nature, despite its more complicated implementation and greater need of computational resources.

The Estimation of Prediction Error

Having constructed a data-based estimation rule, perhaps a logistic regression or a classification tree, the statistician would like to know its performance as a predictor of future cases. There are two main theories concerning prediction error: (1) penalty methods such as Cp, Akaike's information criterion, and Stein's unbiased risk estimate that depend on the covariance between data points and their corresponding predictions; and (2) cross-validation and related nonparametric bootstrap techniques. This article concerns the connection between the two theories. A Rao–Blackwell type of relation is derived in which nonparametric methods such as cross-validation are seen to be randomized versions of their covariance penalty counterparts. The model-based penalty methods offer substantially better accuracy, assuming that the model is believable.

Bootstrap techniques in semiparametric estimation methods for ARFIMA models: A comparison study

This paper considers different bootstrap procedures for investigating the estimation of the fractional parameter d in a particular case of long memory processes, i.e. for ARFIMA models withd in (0.0, 0.5). We propose two bootstrap techniques to deal with semiparametric estimation methods of d. One approach consists of the local bootstrap method for time frequency initially suggested for the ARMA case by Paparoditis and Politis (1999), and the other consists of the bootstrapping in the residuals of the frequency-domain regression equation. Through Monte Carlo simulation, these alternative bootstrap methods are compared, based on the mean and the mean square error of the estimators, with the well-known parametric and nonparametric bootstrap techniques for time series models.

Geriatric performance on the Neurobehavioral Cognitive Status Examination (Cognistat): What is normal?

The Neurobehavioral Cognitive Status Examination (Cognistat) is a widely used cognitive screening measure that has been utilized in several clinical studies with a geriatric population. However, there has been relatively little normative research since its original publication. The objective of this study was to develop age-corrected norms for Cognistat. One hundred and fifty healthy participants aged 60–85 were recruited, all of whom spoke English fluently. Twenty-seven peoples met at least one exclusionary criterion and were therefore excluded. The participants were administered Cognistat along with measures of IQ, depression, alcohol usage, and activity level. Data were not normally distributed; therefore, analysis of these data was completed using descriptive statistics and the nonparametric bootstrapping technique. Study results provide age-corrected profiles that differ significantly from and extend the originally published norms.

Geriatric performance on the Neurobehavioral Cognitive Status Examination (Cognistat) What is normal?

The Neurobehavioral Cognitive Status Examination (Cognistat) is a widely used cognitive screening measure that has been utilized in several clinical studies with a geriatric population. However, there has been relatively little normative research since its original publication. The objective of this study was to develop age-corrected norms for Cognistat. One hundred and fifty healthy participants aged 60-85 were recruited, all of whom spoke English fluently. Twenty-seven peoples met at least one exclusionary criterion and were therefore excluded. The participants were administered Cognistat along with measures of IQ, depression, alcohol usage, and activity level. Data were not normally distributed; therefore, analysis of these data was completed using descriptive statistics and the nonparametric bootstrapping technique. Study results provide age-corrected profiles that differ significantly from and extend the originally published norms.

Application and comparison of two statistical trajectory techniques for identification of source regions of atmospheric aerosol species

Two approaches for identification of source locations and preferred transport pathways of atmospheric particulate trace elements and aerosol species are investigated, namely, versions of the potential source contribution function method (PSCF) and the concentration field method (CF). Both methods are based on combining chemical data with calculated air parcel backward trajectories. The two methods are applied to four multi-species multi-annual concentration time series measured at sites in Finland, Norway, and Israel. A non-parametric bootstrap technique is used to estimate the statistical significance of the calculated PSCF values. It is found that the methods agree well with each other and correctly identify known emission sources. Examples of applying the methods are presented, mainly for the site in Finland.

Statistical nonparametric methods for the study of fossil populations

The precise knowledge of the number and nature of the species belonging to a fossil assemblage as well as of the structure of each species (e.g., age, sex) is of great importance in paleontology. Mixture analysis based on the method of maximum likelihood is a modern statistical technique that concerns the problem of samples consisting of several components, the composition of which is not known. Nonparametric bootstrap and jackknife techniques are used to calculate a confidence interval for each estimated parameter (prior probability, mean, standard deviation) of each group. The bootstrap method is also used to evaluate mathematically how many groups are present in a sample. Experimental density smoothing using the kernel method appears to be a better solution than the use of histograms for the estimation of a distribution. This paper presents some basic concepts and procedures and discusses some preliminary results concerning sex ratios and mortality profile assessments using bones and tooth metric data of small (Ovis antiqua) and large (Bos primigenius) bovines from European Pleistocene sites.