-
[show abstract]
[hide abstract]
ABSTRACT: It is uncertain which methods for the diagnosis of rectal gonococcal and chlamydial infection are optimal. This study evaluated the performance of culture and nucleic acid amplification tests (NAATs) for rectal chlamydial and gonococcal diagnosis. From July 2003 until February 2007, 441 rectal test sets were collected from individuals attending a sexually transmitted disease clinic and three HIV clinics who gave a history of anal intercourse or were women at high risk for Neisseria gonorrhoeae or Chlamydia trachomatis infections. Rectal swab specimens were tested using culture and commercial NAATs employing transcription-mediated amplification (TMA), strand displacement amplification (SDA), and PCR amplification. Test performance was evaluated using a rotating standard by which patients were classified as infected if either two or three comparator tests were positive. Test sensitivities for the detection of N. gonorrhoeae ranged from 66.7% to 71.9% for culture to 100% for TMA. Specificities were 99.7% to 100% for culture and greater than 95.5% for all three NAATs. Test sensitivities for C. trachomatis ranged from 36.1% to 45.7% for culture and among NAATS from 91.4% to 95.8% for PCR to 100% for TMA. Specificities of the NAATs ranged from 95.6% to 98.5% (two-of-three standard) and from 88.8% to 91.8% (three-of-three standard). Over 60% and 80% of gonococcal and chlamydial infections, respectively, among men who have sex with men and over 20% of chlamydial infections in women would have been missed if the rectal site had not been tested. Currently available NAATs are more sensitive for the detection of chlamydial and gonococcal infection at the rectal site than is culture.
Journal of clinical microbiology 03/2010; 48(5):1827-32. · 4.16 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The purpose of this study was to identify mortality risk factors in children with congenital diaphragmatic hernia (CDH) treated with extracorporeal membrane oxygenation (ECMO) and generate a prediction score for those at a very high risk for mortality.
Data on first ECMO runs of all neonates with CDH, between January 1997 and June 2007, were obtained from the Extracorporeal Life Support Organization registry (N = 2678). The data were split into "training data (TD)" (n = 2006) and "validation data" (n = 672). The primary outcome analyzed was in-hospital mortality. Modified Poisson regression was used for analyses.
Overall in-hospital mortality among 2678 neonates (males, 57%; median age at ECMO, 1 day) was 52%. The univariate and multivariable analyses were performed using TD. An empirically weighted mortality prediction score was generated with possible scores ranging from 0 to 35 points. Of 69 who scored 14 or higher in the TD, 62 died (positive predictive value [PPV], 90%), of 37 with 15 or higher, 35 died (PPV, 95%), of 23 with 16 or higher, 22 died (PPV, 96%). A cut-off point of 15 was chosen and was tested using the separate validation dataset. In validation data, the cut-off point 15 had a PPV of 96% (23 died of 24).
Scoring 15 or higher on the prediction score identifies neonates with CDH at a very high risk for mortality among those managed with ECMO and could be used in surgical decision making and counseling.
Journal of Pediatric Surgery 02/2009; 44(1):87-93. · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The optimal methods for the diagnosis of pharyngeal Neisseria gonorrhoeae infection are uncertain. The objective of this study was to define the performance of culture and nucleic acid amplification tests (NAATs) for the diagnosis of pharyngeal N. gonorrhoeae. In this cross-sectional study, males and females >15 years old who acknowledged performing fellatio or cunnilingus (in the previous 2 months) were recruited from three clinics (two human immunodeficiency virus clinics and one sexually transmitted diseases clinic) located in Birmingham, AL. The test performance of culture for N. gonorrhoeae, the Gen-Probe Aptima Combo 2 transcription-mediated amplification assay (TMA), the BD ProbeTec ET amplified DNA strand displacement assay (SDA), and the Roche Cobas Amplicor PCR was defined by using a rotating "gold standard" of any positive results by two or three of the three tests that excluded the test being evaluated. A total of 961 evaluable test sets were collected. On the basis of a rotating gold standard of positive results by two of three comparator tests, the sensitivity and the specificity were as follows: culture for N. gonorrhoeae, 50.0% and 99.4%, respectively; PCR, 80.3% and 73.0%, respectively; TMA, 83.6% and 98.6%, respectively; and SDA, 93.2% and 96.3%, respectively. On the basis of a rotating gold standard of positive results by three of three comparator tests, the sensitivity and specificity were as follows: culture for N. gonorrhoeae, 65.4% and 99.0%, respectively; PCR, 91.9% and 71.8%, respectively; TMA, 100% and 96.2%, respectively; and SDA, 97.1% and 94.2%, respectively. In conclusion, currently available NAATs are more sensitive than culture for the detection of pharyngeal gonorrhea in at-risk patients. PCR is substantially less specific than culture, TMA, or SDA and should not be used for the detection of pharyngeal gonorrhea.
Journal of clinical microbiology 02/2009; 47(4):902-7. · 4.16 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Administrative claims data have a limited ability to identify persons with high compliance to oral bisphosphonates. We tested whether adding information on compliance with other drugs used to treat chronic, asymptomatic conditions would improve the predictive ability of administrative data to identify adherent individuals.
Using data from a large, US healthcare organization, we identified new bisphosphonate users and their 1-year compliance to oral bisphosphonates, quantified by the Medication Possession Ratio (MPR). Multivariable logistic regression models evaluated the relationship between high bisphosphonate compliance (MPR >or=80%) and patient demographics, comorbidities, and health services utilization. To these logistic regression models, we evaluated the incremental change in the area under the receiver operator curve (AUC) after adding information regarding compliance with other drug classes. These included antihyperlipidemics (statins), antihypertensives, antidepressants, oral diabetes agents, and glaucoma medications. Results from the logistic regression models were evaluated in parallel using recursive partitioning trees with 10-fold cross-validation.
Among 101,038 new bisphosphonate users, administrative data identified numerous nonmedication factors (eg, age, gender, use of preventive services) significantly associated with high bisphosphonate compliance at 1 year. However, all these factors in aggregate had low discriminant ability to identify persons highly adherent with bisphosphonates (AUC = 0.62). For persons who were new users of >or=1 of the other asymptomatic condition drugs, MPR data on the other drugs substantially improved the prediction of high bisphosphonate compliance. The impact on prediction was largest for concomitant statin users (AUC = 0.70).
Information on compliance with drugs used to treat chronic asymptomatic conditions improves the prediction of compliance with oral bisphosphonates. This information may help identify persons who should receive targeted interventions to promote compliance to osteoporosis medications.
Medical care 02/2009; 47(3):334-41. · 3.24 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Bayesian data mining methods have been used to evaluate drug safety signals from adverse event reporting systems and allow for evaluation of multiple endpoints that are not prespecified. Their adaptation for use with longitudinal data such as administrative claims has not been previously evaluated or validated.
In this pilot study, we evaluated the feasibility of adapting data mining methods using the empirical Bayes Multi-item Gamma Poisson Shrinkage (MGPS) algorithm to longitudinal administrative claims data. The Medicare Current Beneficiary Survey was used to identify a cohort of Medicare enrollees who were exposed to cyclooxygenase selective (coxib) or nonselective nonsteroidal anti-inflammatory drugs (NS-NSAIDs) from 1999 to 2003. Empirical Bayes MGPS algorithm was used to simultaneously evaluate 259 outcomes associated with current use of coxibs versus NS-NSAIDs while adjusting for key covariates and multiple comparisons. For comparison, a parallel analysis used traditional epidemiologic methods to evaluate the relationship between coxib versus NS-NSAID use and acute myocardial infarction, with the goal of establishing the concurrent validity of the data mining approach.
Among 9431 Medicare beneficiaries using NSAIDs and considering all 259 possible outcomes, empirical Bayes MGPS identified an association between current celecoxib use and acute myocardial infarction (Empirical Bayes Geometric Mean ratio 1.91) but not other outcomes. Rofecoxib use was associated with acute cerebrovascular events (Empirical Bayes Geometric Mean ratio 1.85) and several other diagnoses that likely represented indications for the drug. Results from the analyses using traditional epidemiologic methods were similar and indicated that the data mining results were valid.
Bayesian data mining methods seem useful to evaluate drug safety using administrative data. Further work will be needed to extend these findings to different types of drug exposures and to other claims databases.
Medical care 10/2008; 46(9):969-75. · 3.24 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Bone mass measurement (BMM) is useful to identify persons with low bone mass who are at increased risk for fracture. Given the increased emphasis that is being placed on preventive services such as screening for osteoporosis, we evaluated trends in BMM among Medicare beneficiaries. We studied a 5% sample of Medicare beneficiaries >or=65 yr of age in 1999-2005. We identified claims for BMM tests performed in both facility and nonfacility settings, evaluated temporal trends in use of these tests, and described the proportion of tests attributable to each specialty of physicians submitting claims. We also assessed patterns of serial testing among individuals who were tested more than once. Claims data from all years were pooled to describe the proportion of persons in the population ever tested. From 1999 to 2005, use of central DXA increased by approximately 50%, and use of peripheral DXA declined. The greatest increases in central DXA occurred among internists, family practitioners, and gynecologists. In 1999, the proportion of 65-yr-old women tested was 8.4%; this increased to 12.9% in 2005. Corresponding proportions for men were 0.6% and 1.7%, respectively. Between 40% and 73% of persons receiving central DXA were retested, most at approximately 2-yr intervals. Aggregating data across all years for whites and blacks, 30.0% of women and 4.4% of men underwent central DXA at least once. We conclude that, although use of DXA steadily increased from 1999 to 2005, only approximately 30% of women and 4% of men at least 65 yr old had a central DXA study. Given the importance of central DXA to assess the risk of osteoporotic fractures, strategies to increase central DXA use to test at-risk persons are warranted.
Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research 07/2008; 23(7):1061-7. · 6.04 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The relationship between high adherence to oral bisphosphonates and the risk of different types of fractures has not been well studied among adults of different ages. Using claims data from a large U.S. health care organization, we quantified adherence after initiating bisphosphonate therapy using the medication possession ratio (MPR) and identified fractures. Cox proportional hazards models were used to evaluate the rate of fracture among nonadherent persons (MPR < 50%) compared with highly adherent persons (MPR >or= 80%) across several age strata and a variety of types of clinical fractures. In conjunction with fracture incidence rates among the nonadherent, these estimates were used to compute the number needed to treat with high adherence to prevent one fracture, by age and fracture type. Among 101,038 new bisphosphonate users, the proportion of persons with high adherence at 1, 2, and 3 yr was 44%, 39%, and 35%, respectively. Among 65- to 78-yr-old persons with a physician diagnosis of osteoporosis, the crude and adjusted rate of hip fracture among the nonadherent was 1.96 (95% CI, 1.48-2.60) and 1.74 (95% CI, 1.30-2.31), respectively, resulting in a number needed to treat with high adherence to prevent one hip fracture of 107. The impact of high adherence was substantially less for other types of fractures and for younger persons. Analysis of adherence in a non-time-dependent fashion artifactually magnified differences in fracture rates between adherent and nonadherent persons. The antifracture effectiveness associated with high adherence to oral bisphosphonates varied substantially by age and fracture type. These results provide estimates of absolute fracture effectiveness across age subgroups and fracture types that have been minimally evaluated in clinical trials and may be useful for future cost-effectiveness studies.
Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research 05/2008; 23(9):1435-41. · 6.04 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This study evaluated mortality during 1962 through 2003 and cancer incidence during 1995 through 2003 at a tire manufacturing plant.
The mortality study included 3425 men and women, employed for at least one year. Of these, 3069 were eligible for the cancer incidence study.
Employees experienced 390 deaths compared with 608 expected (standardized mortality ratio (SMR)=64; 95% confidence interval (CI)=58-71). Total cancer mortality (123 observed, SMR=75, CI=62-89) and lung cancer mortality (47 observed, SMR=72, CI=53-96) were lower than expected. Hourly white men had small increases in stomach cancer, bladder cancer, and leukemia deaths. During 1995 through 2003, 169 incident cancers were observed compared with 197 expected (SIR=86, 95% CI=74-100). Three mesothelioma cases occurred among hourly white men (SIR=653, CI=135-1907); all were exposed potentially to asbestos before starting at the rubber plant.
Small numbers and limited information on jobs, occupational agents, and lifestyle preclude attribution of observed increases to workplace exposures.
Journal of Occupational and Environmental Medicine 07/2007; 49(6):680-90. · 2.06 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This investigation assessed the validity of estimates of exposure to 1,3-butadiene (BD) developed for a plant included in a study of mortality among synthetic rubber industry workers. The estimates were developed without using historical measurement data and have not been validated previously.
Personal BD measurements came from an exposure-monitoring program initiated in 1977. For each job, we computed the year-specific difference between the BD estimate and the mean of BD measurements. We also computed rank correlation coefficients and calculated the mean, across all measurements, of the difference between the estimate and the measurement.
The mean BD concentration was 5.2 ppm for 4978 measurements and 4.7 ppm for the corresponding estimates. The mean difference between estimates and measurements was -0.50 ppm (standard deviation, 26.5 ppm) overall and ranged from -227.9 to +27.0 ppm among all 306 job/year combinations. Estimates were correlated with measurements for all 306 combinations (rank correlation coefficient, r=0.45, p<0.0001), for 82 combinations pertaining to jobs that were well-defined by a specific set of tasks and typically found in styrene-BD rubber (SBR) plants (r=0.81, p<0.0001), for 70 combinations pertaining to jobs that were well-defined but not typical (r=0.29, p=0.01) and for 92 combinations pertaining to poorly-defined jobs typically found in SBR plants (r=0.56, <0.0001). Estimates were not correlated with measurements for poorly defined jobs not typically found in SBR plants (r=0.01, p=0.93). For well-defined typical SBR jobs with measurement means that were over 7.0 ppm, estimates were consistently lower than measurements.
Possible reasons for differences between estimates and measurements included faulty assumptions used in developing BD estimates, unstable or nonrepresentive measurements and errors in linking measurement data to the job-exposure matrix. Exposure misclassification may have been more severe for subjects from the validation study plant than for subjects from other plants in the mortality study. BD estimates for typical SBR jobs, which comprise most operations at all but one of the plants in the mortality study, appeared to be useful for ranking workers by cumulative exposure. Uncertainty analyses would enhance the utility of the BD exposure estimates for quantitative risk assessment.
Chemico-Biological Interactions 04/2007; 166(1-3):29-43. · 2.46 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Previous research updated the mortality experience of North American synthetic rubber industry workers during the period 1944-1998, determined if leukemia and other cancers were associated with several employment factors and carried out Poisson regression analysis to examine exposure-response associations between estimated exposure to 1,3-butadiene (BD) or other chemicals and cancer. The present study used Cox regression procedures to examine further the exposure-response relationship between several unlagged and lagged, continuous, time-dependent BD exposure indices (BD parts per million (ppm)-years, the total number of exposures to BD concentrations >100 ppm ("peaks") and average intensity of BD) and leukemia, lymphoid neoplasms and myeloid neoplasms. All three BD exposure indices were associated positively with leukemia. Using continuous, untransformed BD ppm-years the regression coefficient (beta) from an analysis that controlled only for age was 2.9 x 10(-4) (p<0.01); the regression coefficient adjusted for all covariates (age, year of birth, race, plant, years since hire and dimethyldithiocarbamate) was similar in magnitude (beta=3.0 x 10(-4), p=0.04). Lagging exposure had minimal impact on the results for leukemia for any of the three BD exposure indices. In models that controlled only for age, lymphoid neoplasms were associated with BD ppm-years and myeloid neoplasms, with BD peaks, but neither trend was statistically significant after adjusting for multiple covariates. The present results support the presence of a causal relationship between high cumulative exposure and high intensity of exposure to BD and leukemia.
Chemico-Biological Interactions 03/2007; 166(1-3):15-24. · 2.46 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: We examined the relation between cancer mortality and time-dependent cumulative exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) estimated from a concentration- and age-dependent kinetic model of elimination, and we estimated incremental cancer risks at age 75. Data from the National Institute for Occupational Safety and Health study of 3,538 workers with occupational exposure to TCDD were analyzed using standardized mortality ratios and Cox regression procedures. Analyses adjusted for potential confounding by age, year of birth, and race and considered exposure lag periods of 0, 10, or 15 years. Other potential confounders including smoking and other occupational exposures were evaluated indirectly. To explore the influence of extreme values of cumulative TCDD ppt-years, we restricted the analysis to observations with exposure below the 95th percentile or used logarithmic (ln) transformed exposure values. We applied penalized smoothing splines to examine variation in the exposure-response relation across the exposure range. TCDD was not statistically significantly associated with cancer mortality using the full data set, regardless of the lag period. When we restricted the analysis to observations with exposure below the 95th percentile, TCDD was associated positively with cancer mortality, particularly when a 15-year lag was applied (untransformed exposure data: regression coefficient , standard error (s.e.) = 1.4 x 10(-6), p < 0.05; ln-transformed exposure data: , s.e. = 2.9 x 10(-2), p < 0.05). The estimated incremental lifetime risk of mortality at age 75 from all cancers was about 6 to more than 10 times lower than previous estimates derived from this cohort using exposure models that did not consider the age and concentration dependence of TCDD elimination.
Risk Analysis 08/2006; 26(4):1059-71. · 2.37 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Incidence studies of occupational factors and cancer in the United States are problematic because the use of population-based registries to identify cases requires development of historical data on subjects' residences and often severely restricts the time period of follow up. This article describes procedures for addressing these challenges.
We used data from studies of cancer incidence and mortality among microelectronics industry employees to assess various methods for developing residential histories and the relative informativeness of the two studies.
We developed residential histories for 98% of 99,229 mortality study subjects. Analyses making alternative assumptions about residential histories yielded standardized incidence ratios varying by at most 6%. Use of postemployment residential histories increased person-years by up to 62% and increased the observed number of cancers by up to 28%. The proportion of mortality study person-years included in the cancer incidence study ranged from 40% to 77% among work activity subcohorts. The number of observed cancer cases in the incidence study was 60% higher than the number of observed cancer deaths in the mortality study.
Assumptions about residential history had little impact on validity. Use of information sources with national coverage to develop residential histories increased the incidence study's precision. Despite geographic and temporal restrictions, incidence studies provide more data than mortality studies on cancers with good survival. However, the potential for selection bias in incidence studies may vary considerably among subcohorts, indicating the need for cautious interpretation of such research.
Annals of Epidemiology 04/2006; 16(3):170-9. · 3.21 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: We evaluated mortality during 1965 to 1999 among 126,836 workers at two semiconductor facilities and one storage device facility.
We compared employees' cause-specific mortality rates with general population rates and examined mortality patterns by facility, duration of employment, time since first employment, and work activity.
Employees had lower-than-expected mortality overall (6579 observed deaths, standardized mortality ratio [SMR] = 65; 95% confidence interval [CI] = 64-67), for all cancers combined (2159 observed, SMR = 78, 95% CI = 75-81) and for other major diseases. Central nervous system cancer was associated with process equipment maintenance at one of the semiconductor facilities (10 observed, SMR = 247, 95% CI = 118-454). Prostate cancer was associated with facilities/laboratories at the storage device facility (18 observed, SMR = 198, (5% CI = 117-313).
Further evaluation of workplace exposures or independent investigations of similar occupational groups may clarify the interpretation of associations observed in this study.
Journal of Occupational and Environmental Medicine 11/2005; 47(10):996-1014. · 2.06 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This exposure assessment was conducted in the first large study of mortality and cancer incidence in semiconductor and storage device manufacturing.
Unique combinations of division, department and job codes and names (DDJ) from work history records were assigned to work groups and exposure categories. Agent exposure matrices assessed differences in potential exposures between groups. Changes in exposure over time were tracked by dividing the production history into manufacturing eras.
Nineteen work groups were developed to capture 310,351 unique DDJs from 1965-1999. Agent exposure matrices contrasted exposure potential to solvents, metals, and work in cleanrooms between groups, and three manufacturing eras were identified for each site.
The work groups, manufacturing eras and agent matrices have been used to classify workers in the study of cancer incidence and mortality.
Journal of Occupational and Environmental Medicine 11/2005; 47(10):983-95. · 2.06 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Recent studies demonstrating a concentration dependence of elimination of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suggest that previous estimates of exposure for occupationally exposed cohorts may have underestimated actual exposure, resulting in a potential overestimate of the carcinogenic potency of TCDD in humans based on the mortality data for these cohorts. Using a database on U.S. chemical manufacturing workers potentially exposed to TCDD compiled by the National Institute for Occupational Safety and Health (NIOSH), we evaluated the impact of using a concentration- and age-dependent elimination model (CADM) (Aylward et al., 2005) on estimates of serum lipid area under the curve (AUC) for the NIOSH cohort. These data were used previously by Steenland et al. (2001) in combination with a first-order elimination model with an 8.7-year half-life to estimate cumulative serum lipid concentration (equivalent to AUC) for these workers for use in cancer dose-response assessment. Serum lipid TCDD measurements taken in 1988 for a subset of the cohort were combined with the NIOSH job exposure matrix and work histories to estimate dose rates per unit of exposure score. We evaluated the effect of choices in regression model (regression on untransformed vs. ln-transformed data and inclusion of a nonzero regression intercept) as well as the impact of choices of elimination models and parameters on estimated AUCs for the cohort. Central estimates for dose rate parameters derived from the serum-sampled subcohort were applied with the elimination models to time-specific exposure scores for the entire cohort to generate AUC estimates for all cohort members. Use of the CADM resulted in improved model fits to the serum sampling data compared to the first-order models. Dose rates varied by a factor of 50 among different combinations of elimination model, parameter sets, and regression models. Use of a CADM results in increases of up to five-fold in AUC estimates for the more highly exposed members of the cohort compared to estimates obtained using the first-order model with 8.7-year half-life. This degree of variation in the AUC estimates for this cohort would affect substantially the cancer potency estimates derived from the mortality data from this cohort. Such variability and uncertainty in the reconstructed serum lipid AUC estimates for this cohort, depending on elimination model, parameter set, and regression model, have not been described previously and are critical components in evaluating the dose-response data from the occupationally exposed populations.
Risk Analysis 09/2005; 25(4):945-56. · 2.37 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This study evaluated cancer incidence and prostate specific antigen (PSA) testing among workers at a plant in Louisiana (LA) that made atrazine and other triazine herbicides. The study covered the time period 1985 through 1997 and included 2045 subjects, of whom 757 worked for the company that owned the plant and 1288 were contract employees. Linkage with a population-based cancer registry and review of death certificates and plant medical records identified cancer cases. Standardized incidence ratios (SIRs) with 95% confidence intervals (CIs) compared subjects' cancer incidence rates with those of a regional general population. Plant medical records provided data on the proportion receiving PSA tests among male company employees. Subjects had 46 observed and 40 expected cases of all cancers combined (SIR = 114, CI = 83-152) and had 11/6.3 prostate cancers (SIR = 175, CI = 87-312). The prostate cancer excess was greater in actively working company employees (5/1.3, SIR = 394, CI = 128-920) than in contract employees or inactive company employees (6/5.0, SIR = 119, CI = 44-260) and was limited to men under 60 years of age. Of the 11 prostate cancer cases, nine were diagnosed at an early clinical stage. From 1993 to 1999, the proportion of male company employees who had at least one PSA test was 86% for those who reached 40 years of age while actively working and was 98% for those who reached 45 years of age. The observed prostate cancer increase may have been due to the frequent PSA testing of actively working company employees. There is no epidemiologic or other information that clearly supports a causal relation between atrazine and prostate cancer.
Journal of Occupational and Environmental Medicine 12/2002; 44(11):1048-58. · 2.06 Impact Factor
