The Effect of a Regimen of Antifungal Cream Use on Episodes of Acute Adenolymphangitis (ADL) among Lymphedema Patients: An Application Using Marginal Structural Models

Episodes of adenolymphangitis (ADL) are a recurrent clinical aspect of lymphatic filariasis (LF) and a risk factor for progression of lymphedema. Inter-digital entry lesions, often found on the web spaces between the toes of those suffering from lymphedema, have been shown to contribute to the occurrence of ADL episodes. Use of antifungal cream on lesions is often promoted as a critical component of lymphedema management. Our objective was to estimate the observed effect of antifungal cream use on ADL episodes according to treatment regimen among a cohort of lymphedema patients enrolled in a morbidity management program. We estimated this effect using marginal structural models for time varying confounding. In this longitudinal study, we estimate that for every one-unit increase in the number of times one was compliant to cream use through 12 months, there was a 23% (RR = 0.77 (0.62, 0.96)) decrease in the number of ADL episodes at 18 months, however the RR’s were not statistically significant at other study time points. Traditionally adjusted models produced a non-significant RR closer to the null at all time points. This is the first study to estimate the effect of a regimen of antifungal cream on the frequency of ADL episodes. This study also highlights the importance of the consideration and proper handling of time-varying confounders in longitudinal observational studies.

Cancer deaths and cases attributable to lifestyle factors and infections in China, 2013

Background

The burden of cancer in China is high, and it is expected to further increase. Information on cancers attributable to potentially modifiable risk factors is essential in planning preventive measures against cancer. We estimated the number and proportion of cancer deaths and cases attributable to ever-smoking, second-hand smoking, alcohol drinking, low fruit/vegetable intake, excess body weight, physical inactivity, and infections in China, using contemporary data from nationally representative surveys and cancer registries.

Methods

The number of cancer deaths and cases in 2013 were obtained from the National Central Cancer Registry of China and data on most exposures were obtained from the China National Nutrition and Health Survey 2002 or 2006 and Global Adult Tobacco Smoking 2010. We used a bootstrap simulation method to calculate the number and proportion of cancer deaths and cases attributable to risk factors and their corresponding 95% confidence intervals (CIs), allowing for uncertainty in data.

Results

Approximately 718 000 (95% CI 702 100-732 200) cancer deaths in men and 283 100 (278 800-288 800) cancer deaths in women were attributable to the studied risk factors, accounting for 52% of all cancer deaths in men and 35% in women. The numbers for incident cancer cases were 952 500 (95% CI 934 200-971 400) in men and 442 700 (437 200-447 900) in women, accounting for 47% of all incident cases in men and 28% in women. The greatest proportions of cancer deaths attributable to risk factors were for smoking (26%), HBV infection (12%), and low fruit/vegetable intake (7%) in men and HBV infection (7%), low fruit/vegetable intake (6%), and second-hand smoking (5%) in women.

Conclusions

Effective public health interventions to eliminate or reduce exposure from these risk factors, notably tobacco control and vaccinations against carcinogenic infections, can have considerable impact on reducing the cancer burden in China.

Test-Retest Reproducibility of Adult-Reported High School Diet Varies among Racially and Ethnically Diverse US Men and Women

Background

Adolescent diet is thought to play an important role in future chronic disease risk. However, few studies have examined the reproducibility of adult-reported adolescent diet, and evidence for possible differences in reproducibility by demographic characteristics is limited.

Objective

We assessed the ability of adults to consistently report past high school diet over a 1-y period and examined differences in reproducibility by selected demographic characteristics.

Methods

By using age-adjusted partial Spearman (ρ) or Pearson (r) correlations, we assessed 1-y reproducibility for 33 line items, 20 food groups, and 2 dietary patterns of high school diet reported in adulthood via a questionnaire completed by 742 participants in the Cancer Prevention Study 3 (CPS-3) Diet Substudy.

Results

Participants' median age was 53 y (range: 31-70 y), 65.2% were women, 59.8% were non-Hispanic white, 24.8% were non-Hispanic black, and 15.4% were Hispanic. The mean Spearman correlation assessing reproducibility across the 33 line items was 0.60 and ranged from 0.44 to 0.72, with no differences in mean correlations by age, sex, race/ethnicity, education, or body mass index (BMI). Reproducibility was similar across food groups (ρ = 0.62; range: 0.44-0.68), with differences by sex, ethnicity, age, or BMI observed for some food groups (e.g., sugar-sweetened beverages). Pearson correlations for the reproducibility of 2 major eating patterns, "fast food" and "whole food," were 0.73 and 0.72, respectively.

Conclusion

These results show good 1-y reproducibility of assessed high school diet, as reported from memory in adulthood, by line item, food group, and dietary pattern, with noted differences by demographic characteristics.

Norovirus in symptomatic and asymptomatic individuals: cytokines and viral shedding

Noroviruses (NoV) are the most common cause of epidemic gastroenteritis world-wide. NoV infections are often asymptomatic, although individuals still shed large amounts of NoV in their stool. Understanding the differences between asymptomatic and symptomatic individuals would help in elucidating mechanisms of NoV pathogenesis. Our goal was to compare the serum cytokine responses and faecal viral RNA titres of asymptomatic and symptomatic NoV-infected individuals. We tested serum samples from infected subjects (n = 26; 19 symptomatic, seven asymptomatic) from two human challenge studies of GI.1 NoV for 16 cytokines. Samples from prechallenge and days 1-4 post-challenge were tested for these cytokines. Cytokine levels were compared to stool NoV RNA titres quantified previously by reverse transcription-polymerase chain reaction (RT-qPCR). While both symptomatic and asymptomatic groups had similar patterns of cytokine responses, the symptomatic group generally exhibited a greater elevation of T helper type 1 (Th1) and Th2 cytokines and IL-8 post-challenge compared to the asymptomatic group (all P < 0·01). Daily viral RNA titre was associated positively with daily IL-6 concentration and negatively with daily IL-12p40 concentration (all P < 0·05). Symptoms were not associated significantly with daily viral RNA titre, duration of viral shedding or cumulative shedding. Symptomatic individuals, compared to asymptomatic, have greater immune system activation, as measured by serum cytokines, but they do not have greater viral burden, as measured by titre and shedding, suggesting that symptoms may be immune-mediated in NoV infection.

Human norovirus infection and the acute serum cytokine response

Noroviruses (NoV) are the most common cause of epidemic gastroenteritis worldwide. The acute immune response to NoV in humans is poorly understood, hindering research on prevention and treatment. To elucidate the acute immune response and test for cytokine predictors of susceptibility to infection, serum samples from two human NoV challenge studies were tested for 16 cytokines. Subjects who became infected (n = 26) were age-matched with subjects who remained uninfected following NoV challenge (n = 26). Samples were tested from prechallenge and days 1-4 post-challenge. Cytokine responses were compared between infected and uninfected groups. Overall, infected individuals exhibited an elevation in T helper type 1 (Th1) and Th2 cytokines, as well as chemokines interleukin (IL)-8 and monocyte chemoattractant protein (MCP-1), compared to uninfected individuals (all P < 0.05). Most cytokines peaked on day 2 post-challenge in infected subjects, and tumour necrosis factor (TNF)-α, IL-8, and IL-10 remained elevated to day 3. The only cytokine elevated significantly among infected subjects to day 4 post-challenge was IL-10 (P = 0.021). Prechallenge cytokine concentrations were not predictive of infection status post-challenge. There were no significant changes in serum cytokines among NoV-challenged subjects who remained uninfected. These results suggest that NoV infection elicits a Th1-type response, with some Th2 activation. Persistent elevation of IL-10 among infected subjects is consistent with activation of adaptive immune responses, such as B cell expansion, as well as down-regulation of Th1 cytokines. This study presents the first comprehensive description of the acute cytokine response to GI.1 NoV in humans.

Characterizing concentrations of diethylene glycol and suspected metabolites in human serum, urine, and cerebrospinal fluid samples from the Panama DEG mass poisoning

CONTEXT

Diethylene glycol (DEG) mass poisoning is a persistent public health problem. Unfortunately, there are no human biological data on DEG and its suspected metabolites in poisoning. If present and associated with poisoning, the evidence for use of traditional therapies such as fomepizole and/or hemodialysis would be much stronger.

OBJECTIVE

To characterize DEG and its metabolites in stored serum, urine, and cerebrospinal fluid (CSF) specimens obtained from human DEG poisoning victims enrolled in a 2006 case-control study.

METHODS

In the 2006 study, biological samples from persons enrolled in a case-control study (42 cases with new-onset, unexplained AKI and 140 age-, sex-, and admission date-matched controls without AKI) were collected and shipped to the Centers for Disease Control and Prevention (CDC) in Atlanta for various analyses and were then frozen in storage. For this study, when sufficient volume of the original specimen remained, the following analytes were quantitatively measured in serum, urine, and CSF: DEG, 2-hydroxyethoxyacetic acid (HEAA), diglycolic acid, ethylene glycol, glycolic acid, and oxalic acid. Analytes were measured using low resolution GC/MS, descriptive statistics calculated and case results compared with controls when appropriate. Specimens were de-identified so previously collected demographic, exposure, and health data were not available. The Wilcoxon Rank Sum test (with exact p-values) and bivariable exact logistic regression were used in SAS v9.2 for data analysis.

RESULTS

The following samples were analyzed: serum, 20 case, and 20 controls; urine, 11 case and 22 controls; and CSF, 11 samples from 10 cases and no controls. Diglycolic acid was detected in all case serum samples (median, 40.7 mcg/mL; range, 22.6-75.2) and no controls, and in all case urine samples (median, 28.7 mcg/mL; range, 14-118.4) and only five (23%) controls (median, < Lower Limit of Quantitation (LLQ); range, < LLQ-43.3 mcg/mL). Significant differences and associations were identified between case status and the following: 1) serum oxalic acid and serum HEAA (both OR = 14.6; 95% C I = 2.8-100.9); 2) serum diglycolic acid and urine diglycolic acid (both OR > 999; exact p < 0.0001); and 3) urinary glycolic acid (OR = 0.057; 95% C I = 0.001-0.55). Two CSF sample results were excluded and two from the same case were averaged, yielding eight samples from eight cases. Diglycolic acid was detected in seven (88%) of case CSF samples (median, 2.03 mcg/mL; range, < LLQ, 7.47).

DISCUSSION

Significantly elevated HEAA (serum) and diglycolic acid (serum and urine) concentrations were identified among cases, which is consistent with animal data. Low urinary glycolic acid concentrations in cases may have been due to concurrent AKI. Although serum glycolic concentrations among cases may have initially increased, further metabolism to oxalic acid may have occurred thereby explaining the similar glycolic acid concentrations in cases and controls. The increased serum oxalic acid concentration results in cases versus controls are consistent with this hypothesis.

CONCLUSION

Diglycolic acid is associated with human DEG poisoning and may be a biomarker for poisoning. These findings add to animal data suggesting a possible role for traditional antidotal therapies. The detection of HEAA and diglycolic acid in the CSF of cases suggests a possible association with signs and symptoms of DEG-associated neurotoxicity. Further work characterizing the pathophysiology of DEG-associated neurotoxicity and the role of traditional toxic alcohol therapies such as fomepizole and hemodialysis is needed.

Physical activity and cardiovascular risk factors at menopause: the Nord-Trøndelag health study

BACKGROUND

Lowered physical activity levels may partially explain changes in metabolic risk factors in women after menopause.

OBJECTIVES

To evaluate the association between physical activity and metabolic risk factors at baseline and after 11 years, as well as the change in that association over time in women who were premenopausal and ≥ 40 years at baseline.

METHODS

Subjects in a Norwegian population-based health survey answered questionnaires and had body and serum measurements during 1995-1997 (HUNT 2) and in a follow-up study during 2006-2008 (HUNT 3). Repeated-measures analyses were used to estimate the association between physical activity and metabolic factors, adjusting for age, smoking status, education, alcohol intake, and parity. Adjustment for hormonal treatment and medication was made, as appropriate.

RESULTS

In women remaining premenopausal, a higher physical activity score in HUNT 3 was associated with lower weight (p < 0.01) and waist-hip ratio (p < 0.01) and higher high density lipoprotein (HDL) cholesterol in HUNT 3 (p < 0.01). In women that were postmenopausal by the time of follow-up, a higher physical activity score in HUNT 3 was associated with lower weight (p < 0.01), waist-hip ratio (p < 0.01), triglycerides (p < 0.01), and higher total cholesterol (p < 0.05), HDL cholesterol (p < 0.01), and diastolic blood pressure (p < 0.05) in HUNT 3. The association of total physical activity score with weight and waist-hip ratio was stronger in HUNT 3 than in HUNT 2 (p < 0.01).

CONCLUSION

Increased physical activity may reduce the risk of adverse outcomes and use of pharmacological management in women of menopausal age.

Comparison of emergency department and hospital admissions data for air pollution time-series studies

BACKGROUND

Emergency department (ED) visit and hospital admissions (HA) data have been an indispensible resource for assessing acute morbidity impacts of air pollution. ED visits and HAs are types of health care visits with similarities, but also potentially important differences. Little previous information is available regarding the impact of health care visit type on observed acute air pollution-health associations from studies conducted for the same location, time period, outcome definitions and model specifications.

METHODS

As part of a broader study of air pollution and health in St. Louis, individual-level ED and HA data were obtained for a 6.5 year period for acute care hospitals in the eight Missouri counties of the St. Louis metropolitan area. Patient demographic characteristics and diagnostic code distributions were compared for four visit types including ED visits, HAs, HAs that came through the ED, and non-elective HAs. Time-series analyses of the relationship between daily ambient ozone and PM₂.₅ and selected cardiorespiratory outcomes were conducted for each visit type.

RESULTS

Our results indicate that, compared with ED patients, HA patients tended to be older, had evidence of greater severity for some outcomes, and had a different mix of specific outcomes. Consideration of 'HA through ED' appeared to more effectively select acute visits than consideration of 'non-elective HA'. While outcomes with the strongest observed temporal associations with air pollutants tended to show strong associations for all visit types, we found some differences in observed associations for ED visits and HAs. For example, risk ratios for the respiratory disease-ozone association were 1.020 for ED visits and 1.004 for 'HA through ED'; risk ratios for the asthma/wheeze-ozone association were 1.069 for ED visits and 1.106 for 'HA through ED'. Several factors (e.g. age) were identified that may be responsible, in part, for the differences in observed associations.

CONCLUSIONS

Demographic and diagnostic differences between visit types may lead to preference for one visit type over another for some questions and populations. The strengths of observed associations with air pollutants sometimes varied between different health care visit types, but the relative strengths of association generally were specific to the pollutant-outcome combination.

Physical activity and age at menopause: the Nord-Trøndelag population-based health study

BACKGROUND

Age at menopause may affect women's subsequent morbidity and mortality. In contrast to numerous other health outcomes, little is known about the possible effects of physical activity on age at menopause.

OBJECTIVES

To assess the relationship between leisure-time physical activity and age at menopause.

METHODS

Premenopausal women participating in a population-based health survey (HUNT 2) conducted in the county of Nord-Trøndelag, Norway reported their physical activity in the period of 1995-1997. Age at menopause was reported during 2006-2008 (HUNT 3). Cox proportional hazards models were used to estimate hazard ratios for menopause and logistic regression to estimate odds ratios for early menopause, with 95% confidence intervals, adjusting for age at menarche, parity, use of oral contraceptives prior to the 6 months preceding participation in HUNT 2, symptoms of depression, smoking status, and education.

RESULTS

Women aged 40-49 years at baseline had lower hazard ratios for menopause when participating in any light leisure-time physical activity compared with no activity (p < 0.05) and similar results were observed in 19-39-year-olds. In 50-59-year-old women, the results varied greatly and did not reach statistical significance.

CONCLUSIONS

The effects of leisure-time physical activity on age at menopause may be age-dependent. We found indications of earlier menopause for the least active women aged 19-49 years at baseline.

Race, neighbourhood characteristics and disparities in chemotherapy for colorectal cancer

BACKGROUND

Studies have found significant race/ethnic and age differences in receipt of adjuvant chemotherapy for stages III colon and II/III rectal cancers. Little is known about the role of neighbourhood factors in these disparities.

METHODS

The 4748 Black and White patients from the Georgia Comprehensive Cancer Registry were diagnosed with stages III colon and II/III rectal cancers between 2000 and 2004. Neighbourhood poverty, segregation (% Black residents) and rurality were linked to each patient using census tract identifiers. Multilevel analyses explored the role of neighbourhood characteristics and the nested association of patient race within categories of neighbourhoods in receipt of chemotherapy.

RESULTS

Odds of receiving chemotherapy for urban and suburban patients were 38% (95% CI 1.09 to 1.74) and 53% (95% CI 1.20 to 1.94) higher than for rural patients. However, odds of receiving chemotherapy for urban Black patients were 24% (95% CI 0.62 to 0.94) lower than for their White counterparts. Receipt of chemotherapy did not significantly differ between Blacks and Whites residing in suburban or rural areas.

CONCLUSION

Black-White disparities in receipt of chemotherapy among Georgia colorectal cancer patients were confined to urban patients. Disparities in receipt of this treatment for rural patients were found irrespective of patient race. Our findings highlight geographic areas where targeted interventions might be needed.

Physical activity and fertility in women: the North-Trøndelag Health Study

BACKGROUND

Changes in the state of energy balance owing to changes in physical activity may affect the reproductive system. We evaluated the association between physical activity (PA) and fertility and parity in healthy women.

METHODS

A population-based health survey (HUNT 1) was conducted during 1984-1986 in Nord-Trøndelag county, Norway, with follow-up from 1995 to 1997 (HUNT 2). The study included 3887 women, <45 years old in HUNT 2. PA was assessed by baseline questionnaire, and fertility and parity by questionnaire at follow-up. Data focused on overall occurrence of infertility in the population (without biological confirmation).

RESULTS

Increased frequency, duration and intensity of PA were associated with increased subfertility, and frequency of PA was associated with voluntary childlessness (P < 0.01). After adjusting for age, parity, smoking, and marital status, women who were active on most days were 3.2 times more likely to have fertility problems than inactive women. Exercising to exhaustion was associated with 2.3 times the odds of fertility problems versus low intensity. Women with highest intensity of PA at baseline had the lowest frequency of continuing nulliparity and highest frequency of having three or more children during follow-up (P < 0.05). Sensitivity analysis including body mass index as confounder did not alter the results. No associations were found between lower activity levels and fertility or parity.

CONCLUSION

Increased risk of infertility was only found for the small group of women reporting the highest levels of intensity and frequency of PA. Awareness of the possible risks of infertility should be highlighted among non-athletic women who exercise vigorously.

Hospital characteristics associated with surgery for non-small cell lung cancer

6543

Background: Previous research suggests that black patients are less likely to undergo curative-intent surgery for early stage non-small cell lung cancer (NSCLC) compared to whites, holding all else constant. Among Medicare beneficiaries 65 and older, the likelihood of patients undergoing surgery is significantly reduced in hospitals with racial compositions of > 30% black patients after controlling for hospital and patient characteristics. This study analyzes whether hospital racial composition is associated with curative-intent surgery among early stage NSCLC patients of all ages. Methods: All adult patients ages 19–104 diagnosed with an invasive initial primary early stage (TNM I-II) NSCLC during 2003–2005 were selected from the National Cancer Data Base (NCDB). Facility characteristics were extracted from the NCDB and American College of Surgeons (ACoS) Commission on Cancer (CoC) Facility Information Profile System (FIPS). Hospital racial composition of lung cancer patients, operationalized as percent black, was initially divided into tertiles; sensitivity analyses used dichotomous definitions. Generalized estimating equations with a logistic model were used to control for clustering by facility. Results: Of 52,853 evaluable patients seen at CoC-approved hospitals, blacks were primarily (74%) seen at hospitals with a racial composition of > 11% black patients. Hispanics (72%), non-Hispanic whites (69%), and Asians or other (72%) were primarily seen at hospitals with < 11% black patients. In sensitivity analyses, irrespective of the cutpoint used to classify hospitals by racial composition of black patients (10, 20, or 30%), black patients seen at hospitals with a high black racial composition were significantly less likely to undergo curative-intent surgery than black patients seen at hospitals with a lower black racial composition. After controlling for patient characteristics in multivariate models, however, high black racial composition at the facility level was no longer significant. Conclusions: Hospital racial composition (% black) was not predictive of reduced likelihood of curative-intent surgery among patients > 18 years of age with NSCLC. Patient-level predictors accounted for the majority of the variation in likelihood of curative-intent surgery.

No significant financial relationships to disclose.

The issue of confounding in epidemiological studies of ambient air pollution and pregnancy outcomes

BACKGROUND

Relationships between ambient air pollution levels during pregnancy and adverse pregnancy outcomes have been investigated using one of three analytic approaches: ambient pollution levels have been contrasted over space, time or both space and time. Although the three approaches share a common goal, to estimate the causal effects of pollution on pregnancy outcomes, they face different challenges with respect to confounding.

METHODS

A framework based on counterfactual effect definitions to examine issues related to confounding in spatial, temporal, and spatial-temporal analyses of air pollution and pregnancy outcomes is presented, and their implications for inference are discussed.

RESULTS

In spatial analyses, risk factors that are spatially correlated with pollution levels are confounders; the primary challenges relate to the availability and validity of risk factor measurements. In temporal analyses, where smooth functions of time are commonly used to control for confounding, concerns relate to the adequacy of control and the possibility that abrupt changes in risk might be systematically related to pollution levels. Spatial-temporal approaches are subject to challenges faced in both spatial and temporal analyses.

CONCLUSION

Each approach faces different challenges with respect to the likely sources of confounding and the ability to control for that confounding because of differences in the type, availability, and quality of information required. Thoughtful consideration of these differences should help investigators select the analytic approach that best promotes the validity of their research.

Maternal plasma cytokines in early- and mid-gestation of normal human pregnancy and their association with maternal factors

Few studies have assessed longitudinal changes in circulating cytokine levels during normal pregnancy. We have examined the natural history of maternal plasma cytokines from early- to mid-pregnancy in a large, longitudinal cohort. Multiplex flow cytometry was used to measure interleukin (IL)-2, IL-6, IL-12, tumor necrosis factor (TNF)-alpha, interferon (IFN)-gamma and granulocyte-macrophage colony-stimulating factor (GM-CSF) in early- (median [IQR]: 8.5 weeks [7.1, 10.0]) and mid-pregnancy (25.0 [24.1, 26.1]) from 1274 Danish women delivering singleton term infants. GM-CSF decreased from early- to mid-pregnancy (median percent change [95% CI]: -51.3% [-59.1%, -41.8%]), while increases were observed in IL-6 (24.3% [4.6%, 43.9%]), IL-12 (21.3% [8.9%, 35.7%]) and IFN-gamma (131.7% [100.2%, 171.6%]); IL-2 (-2.8% [-11.5%, 0.0%]) and TNF-alpha (0% [-5.9%, 25.6%]) remained stable. Positive correlations were found between all cytokines, both in early- and mid-pregnancy (all p<0.001). Early- and mid-pregnancy levels were rank-correlated for IL-2, IL-12, TNF-alpha and GM-CSF, but not IL-6 and IFN-gamma; these correlations were generally weaker than correlations between different cytokines at a single time point in pregnancy. Women with a pre-pregnancy BMI <18.5 had reduced levels of IFN-gamma and GM-CSF compared to women in other BMI categories, while women aged >or=35 years had elevated IL-2, IL-6, TNF-alpha and IFN-gamma. Early-pregnancy levels of TNF-alpha were higher in women with a prior preterm delivery. Cytokine levels were not associated with gravidity. In conclusion, cytokines were detected in plasma during early- and mid-pregnancy, with IL-6, IL-12, IFN-gamma and GM-CSF concentrations varying over pregnancy. Concentrations may depend on BMI, maternal age and prior preterm delivery.

Blood lead levels and risk factors for lead poisoning in children and caregivers in Chuuk State, Micronesia

Lead poisoning is a preventable environmental disease. Children and developing fetuses are especially vulnerable; even low blood lead levels (BLLs) are linked with learning and behavioral problems. We assessed children's and their caregivers' BLLs and risk factors for lead exposure in Chuuk State, Federated States of Micronesia. Children aged 2-6 years were randomly selected within 20 randomly selected villages. Children and caregivers provided venous blood, and caregivers offered information about possible risk factors for lead exposure. Mean BLLs were 39 microg/l for children and 16 microg/l for caregivers. Children with BLLs of > or = 100 microg/l (elevated) were 22.9 (95% CI: 4.5-116.0) times more likely to have a caregiver with an elevated BLL, 6.2 (95% CI: 1.4-27.3) times more likely to live on an outer island, and 3.4 (95% CI: 1.7-6.9) times more likely to have a family member who made lead fishing weights than did other children even after controlling for age and sex. For children, 61% of elevated BLLs could be attributed to making fishing weights. Caregivers with elevated BLLs were 5.9 (95% CI: 1.5-23.7) times more likely to live in a household that melted batteries than other caregivers even after controlling for age and education. For caregivers, 37% of the elevated BLLs could be attributed to melting batteries. The association of elevated BLLs in children and their caregiver suggests a common environmental exposure. Melting batteries to make fishing sinkers is a preventable source of lead exposure for children and their caregivers in Chuuk. Published by Elsevier GmbH.

Tests of trait—haplotype association when linkage phase is ambiguous, appropriate for matched case-control and cohort studies with competing risks

The impact of competing risks on tests of association between disease and haplotypes has been largely ignored. We consider situations in which linkage phase is ambiguous and show that tests for disease-haplotype association can lead to rejection of the null hypothesis, even when true, with more than the nominal 5 per cent frequency. This problem tends to occur if a haplotype is associated with overall mortality, even if the haplotype is not associated with disease risk. A small simulation study illustrates the magnitude of bias (high type I error rate) in the context of a cohort study in which a modest number of disease cases (about 350) occur over time. The bias remains even if the score test is based on a logistic model that includes age as a covariate. For cohort studies, we propose a new test based on a modification of the proportional hazards model and for case-control studies, a test based on a conditional likelihood that have the correct size under the null even in the presence of competing risks, and that can be used when haplotype is ambiguous.

Is readmission to hospital an indicator of poor process of care for patients with heart failure?

BACKGROUND

Controversy exists about the appropriateness of using readmission as an indicator of the quality of care. A study was undertaken to measure the validity and predictive ability of readmission in this context.

METHODS

An evaluation study was performed in patients discharged alive with heart failure from three Swiss academic medical centres. Process quality indicators were derived from evidence based guidelines for the management and treatment of heart failure. Readmissions were calculated from hospital administrative data. The predictive ability of readmissions was evaluated using bivariate and multivariate analyses, and validity by calculating sensitivity, specificity, positive and negative predictive value, using process indicators as the "gold standard".

RESULTS

Of 1055 eligible patients discharged alive, 139 (13.2%) were readmitted within 30 days. The adjusted odds ratio (OR) for absence of measurement of left ventricular function was 0.70 (95% CI 0.45 to 1.08) for readmissions. In patients with left ventricular systolic dysfunction, three dose categories of angiotensin converting enzyme inhibitor were examined using ordinal logistic regression. The adjusted OR for these categories was 1.07 (95% CI 0.56 to 2.06) for readmissions. When using process indicators as the gold standard to assess the validity of readmissions, sensitivity ranged from 0.08 to 0.17 and specificity from 0.86 to 0.93.

CONCLUSIONS

Readmission did not predict and was not a valid indicator of the quality of care for patients with heart failure admitted to three Swiss university hospitals.

Assessment of two self-reporting questionnaires for eating disorders in women

OBJECTIVE

The goals of this study were to evaluate how well the Eating Disorder Inventory (EDI) predicted specific eating disorders as identified by the Survey of Eating Disorders (SED), and to identify the most predictive EDI scales. An additional goal was to evaluate the EDI as a screening instrument for eating disorders among a large non-clinical group of women.

METHOD

The community-based study sample consisted of 1245 Norwegian women between age 18 and 50 years (mean 23). Subjects filled out the EDI and the SED.

RESULTS

The strongest predictors of SED-defined eating disorders were the EDI scales bulimia (BN) and drive of thinness (DT) and body mass index (BMI). When we classified women with an EDI sum score of 40 or higher as having an eating disorder (ED), the sensitivity was 38% and the specificity was 90%, compared to the SED diagnosis of ED.

DISCUSSION

Our findings suggest modest agreement between the EDI and the SED, and modest sensitivity and specificity of the EDI. While these instruments, perhaps with some modification, may be adequate for use in selected research applications, they should probably be supplemented with additional information in a clinical or treatment setting.

Eating disorder behavior in physically active Norwegian women

Competitive female athletes may have higher risk of eating disorders (ED) than recreational athletes, in part because low body fat may be advantageous in sports. However, women who participate in sports or physical activities might also do this for weight reduction, altering body shape, and for other food related reasons. This study was conducted to evaluate the association between hours of leisure time physical activity (PA) and prevalence of ED, and whether the association between the Eating Disorders Inventory (EDI) scales and ED, as classified by the Survey of Eating Disorders (SED), was the same among women with high vs. low physical activity levels. The community-based study sample consisted of 898 female students aged 18-50 years. Subjects filled out the EDI, the SED, and a questionnaire concerning physical activity and demographic data. Physical activity was not associated with increased risk of having a SED-defined diagnosis of an ED or of having an EDI score greater than 40, which is sometimes used to define women at risk for having an eating disorder. Women with a SED-defined ED had higher scores on all subscales of the EDI, but did not differ with respect to age, BMI or hours of weekly activity. The strongest predictors of having a SED-defined ED were body mass index (BMI), two EDI scales, drive of thinness and Bulimia Nervosa, with no differences between the two PA groups. Among women with a SED-defined ED, those with more than 5 h weekly PA did not differ from others with respect to scoring on the EDI or BMI. Results suggest that hours of physical activity may not increase risk of developing ED. Effective interventions are needed to help the growing numbers of persons with body-image and eating difficulties. For some, moderate physical training might be helpful.

New estimator of the genotype risk ratio for use in case-parental control studies

Estimation of the genotype risk ratio can be an important part of studying the role of genetics in disease causation. For example, one might estimate risk among persons with genotype DD compared with risk among those with genotype Dd, where the candidate locus has alleles D and d, with D representing the disease susceptibility allele. In this paper, the authors propose a modified method of analysis for case-parental control studies that can improve efficiency. They show how investigators can use information from families in which both parents are observed to improve the estimator created by Sun et al., which applies when only one parent and an affected offspring have been observed. Since this information is not used by the conditional approach of Schaid and Sommer, the authors' approach allows for more complete use of available information, leading to a smaller mean squared error of the genotype risk ratio estimators. The authors also suggest a way to combine estimates from families in which one parent and one offspring are observed and estimates from families in which both parents and one offspring are observed.

Statistical models to predict the need for postoperative intensive care and hospitalization in pediatric surgical patients

OBJECTIVE

To develop statistical models for predicting postoperative hospital and ICU stay in pediatric surgical patients based on preoperative clinical characteristics and operative factors related to the degree of surgical stress. We hypothesized that preoperative and operative factors will predict the need for ICU admission and may be used to forecast the length of ICU stay or postoperative hospital stay.

DESIGN

Prospective data collection from 1,763 patients.

SETTING

Tertiary care children's hospital.

PATIENTS AND PARTICIPANTS

All pediatric surgical patients, including those undergoing day surgery. Patients undergoing dental or ophthalmologic surgical procedures were excluded.

INTERVENTIONS

None.

MEASUREMENTS AND RESULTS

A logistic regression model predicting ICU admission was developed from all patients. Poissonregression models were developed from 1,161 randomly selected patients and validated from the remaining 602 patients. The logistic regression model for ICU admission was highlypredictive (area under the receiver operating characteristics (ROC) curve = 0.981). In the data set used for development of Poisson regression models, significant correlations occurred between the observed and predicted ICU stay (Pearson r = 0.468, p < 0.0001, n = 131) and between the observed and predicted hospital stay for patients undergoing general (r = 0.695, p < 0.0001), orthopedic (r = 0.717, p < 0.0001), cardiothoracic (r = 0.746, p < 0.0001), urologic (r = 0.458, p < 0.0001), otorhinolaryngologic (r = 0.962, p < 0.0001), neurosurgical (r = 0.7084, p < 0.0001) and plastic surgical (r = 0.854, p < 0.0001) procedures. In the validation data set, correlations between predicted and observed hospital stay were significant for general (p < 0.0001), orthopedic (p < 0.0001), cardiothoracic (p = 0.0321) and urologic surgery (p = 0.0383). The Poisson models for length of ICU stay, otorhinolaryngology, neurosurgery or plastic surgery could not be validated because of small numbers of patients.

CONCLUSIONS

Preoperative and operative factors may be used to develop statistical models predicting the need for ICU admission in pediatric surgical patients, and hospital stay following general surgical, orthopedic, cardiothoracic and urologic procedures. These statistical models need to be refined and validatedfurther, perhaps using data collection from multiple institutions.

Accounting for unmeasured population substructure in case-control studies of genetic association using a novel latent-class model

We propose a novel latent-class approach to detect and account for population stratification in a case-control study of association between a candidate gene and a disease. In our approach, population substructure is detected and accounted for using data on additional loci that are in linkage equilibrium within subpopulations but have alleles that vary in frequency between subpopulations. We have tested our approach using simulated data based on allele frequencies in 12 short tandem repeat (STR) loci in four populations in Argentina.

Transmission/disequilibrium tests for quantitative traits

The transmission/disequilibrium test (TDT) is a powerful method of locating disease genes. The TDT was originally proposed for use in studies of qualitative traits in families with both parents available. Recently, the TDT has been extended to studies of qualitative traits in sibships without parents available and in families with one parent available. It has also been extended for use in studies of quantitative traits in families with both parents available and in sibships with multiple offspring. In this paper, we first propose a new class of TDT-type tests for linkage in the presence of linkage disequilibrium for use in studies of families with both parents available. The TDT of Spielman et al. (1993) for qualitative traits and the TDT of Rabinowitz (1997) for quantitative traits are special cases of the new tests. Second, we propose a new class of TDT-type tests for linkage for use in studies of families with one parent available. Third, we study the validity and the power of the tests using simulations. Finally, we propose a method of combining data from different types of families. The combined test is valuable and allows researchers full use of the available data in detecting linkage between a marker locus and an unobservable quantitative trait locus. An important feature of the tests proposed in this paper is that no assumptions on the distribution of the quantitative traits are needed.

Interim results of the study of particulates and health in Atlanta (SOPHIA)

Substantial evidence supports an association of particulate matter (PM) with cardiorespiratory illnesses, but little is known regarding characteristics of PM that might contribute to this association and the mechanisms of action. The Atlanta superstation sponsored by the Electric Power Research Institute as part of the Aerosol Research and Inhalation Epidemiology Study (ARIES) study is monitoring chemical composition of ambient particles by size fraction, as well as a comprehensive suite of other pollutants, at a site in downtown Atlanta during the 25-month period, August 1, 1998-August 31, 2000. Our investigative team is making use of this unique resource in several morbidity studies, called the "Study of Particulates and Health in Atlanta (SOPHIA)". The study includes the following components: (1) a time series investigation of emergency department (ED) visits for the period during which the superstation is operating; (2) a time series investigation of ED visits during the 5 years prior to implementation of the superstation; and (3) a study of arrhythmic events in patients equipped with automatic implantable cardioverter defibrillators (AICDs) for the period January 1, 1993-August 31, 2000. Thirty-three of 39 Atlanta area EDs are participating in the ED studies, comprising over a million annual ED visits. In this paper, we present initial analyses of data from 18 of the 33 participating EDs. The preliminary data set includes 1,662,713 ED visits during the pre-superstation time period and 559,480 visits during the superstation time period. Visits for four case groupings--asthma, chronic obstructive pulmonary disease (COPD), dysrhythmia, and all cardiovascular diseases (CVDs) combined--have been assessed relative to daily air quality indices, controlling for long-term temporal trends and meteorologic variables, using general linear models, generalized estimating equations and generalized additive models. Single-pollutant models predicting case visitation rates using moving averages of 0-, 1-, and 2-day lagged air quality variables were run. For the pre-superstation period, PM10 (24-h), ozone (8-h), SO2 (1-h), NO2 (1-h) and CO (1-h) were studied. For the first 12 months of superstation operation, the following air quality variables of a priori interest were available: ozone (8-h), NO2 (1-h), SO2 (1-h), CO (1-h), and 24-h measurements of PM10, coarse PM (PM 2.5-10 microm), PM2.5, polar VOCs, 10-100 nm particulate count and surface area, and in the PM2.5 fraction: sulfates, acidity, water-soluble metals, organic matter (OM), and elemental carbon (EC). During the pre-superstation time period, statistically significant, positive associations were observed for adult asthma with ozone, and for COPD with ozone, NO2 and PM10. During the superstation time period, the following statistically significant, positive associations were observed: dysrhythmia with CO, coarse PM, and PM2.5 EC; and all CVDs with CO, PM2.5 EC and PM2.5 OM. While covariation of many of the air quality indices limits the informativeness of this analysis, the study provides one of the first assessments of PM components in relation to ED visits.

Predictors of correct treatment of children with fever seen at outpatient health facilities in the Central African Republic

To identify factors associated with improved performance of health care workers who treat ill children in developing countries, the authors analyzed a sample of consultations of children with malaria (defined as any fever) from a national health facility survey conducted in the Central African Republic from December 1995 to January 1996. Twenty-eight health care workers and 204 children were studied. A univariate analysis revealed the following significant predictors of correct treatment, as defined by the Central African malaria control program: high fever (odds ratio (OR) = 3.25, 95% confidence interval (CI): 1.47, 7.17); correct health care worker diagnosis (OR = 2.59, 95% CI: 1.39, 4.85); and the caregiver's reporting the child's fever to the health care worker (OR = 2.18, 95% CI: 1.32, 3.62). There was an unexpected inverse association between the presence of a fever treatment chart and correct treatment (OR = 0.19, 95% CI: 0.04, 0.91). Correct treatment was marginally associated with a longer consultation time (p value for trend = 0.058). Neither in-service training in the treatment of fever nor supervision was significantly associated with correct treatment. For child health programs to improve, targeted studies are needed to understand which factors, alone or in combination, improve health care worker performance.

On the use of population-based registries in the clinical validation of genetic tests for disease susceptibility

PURPOSE

Many new genetic tests for susceptibility to adult-onset diseases are developed on the basis of selected and high-risk groups. Before such tests can be used in medical practice, however, epidemiologic studies must be conducted to evaluate their clinical sensitivity, specificity, and positive predictive value in the general population. For many common adult-onset diseases, this process may take decades of follow-up.

METHOD

We illustrate how clinical validation of new predictive genetic tests can be done retrospectively using case-control studies that are derived from population-based registries of diseases. We use the examples of birth defects and cancer registries to illustrate a hypothetical process by which such tests can be clinically validated.

RESULTS

We demonstrate how such epidemiologic studies can be successfully used to derive measures of a test's sensitivity, specificity, positive predictive value, negative predictive value, and of the population attributable fraction of disease due to the disease-susceptibility genes. Under certain assumptions, data derived from population-based case-control studies provide adequate estimates of lifetime risks for disease (penetrance) among people with specified genotypes.

CONCLUSIONS

With adequate protections of human subjects, studies involving population-based registries of disease will increasingly become valuable in validating the numerous genetic tests that will emerge from advances in human genetic research and the Human Genome Project.

Bias associated with study protocols in epidemiologic studies of disease familial aggregation

The effect of selection bias has not been well evaluated in epidemiologic studies which focus on familial aggregation. The authors illustrate this type of bias for a reconstructed cohort study. With the reconstructed cohort design, cases and controls are first selected from the population and their relatives form the exposed and unexposed cohorts, respectively. The recurrence risk ratio (RRR) is calculated to assess and measure familial aggregation. The ways of utilizing information from relatives affects the estimate of RRR, and the authors show that a traditional method used in epidemiologic studies can yield a severely biased estimate of the RRR. However, this traditional approach can give approximately unbiased estimates under special conditions. A novel selection approach is proposed which yields an unbiased estimate of RRR. In conclusion, when relatives are identified through cases or controls, they should be included and counted in the study cohorts each time a case or control is selected, even if they or other family members have already been included.

A model to estimate the probability of hepatitis B- and Haemophilus influenzae type b-vaccine uptake into national vaccination programs

Most countries have been slow to adopt new vaccines into national vaccination schedules, despite recommendations from global multi-lateral agencies. Characteristics of countries that have adopted hepatitis B (HB) vaccine were analysed and used to formulate a logistic regression model. The model was applied to country-specific data to predict HB and Haemophilus influenzae type b (Hib) vaccine uptake. The greatest predictors of HB uptake were coverage rates of other vaccines, vaccine cost relative to the economy, and perceived disease burden. The logistic regression model's probability estimate of vaccine uptake agreed well with observed data for HB and Hib, (c-statistic 85 and 82%, respectively). Application of this model to other antigens may aid in predicting potential national markets to better plan new vaccine supply and demand.

Occurrence of primary cancers in association with multiple myeloma and Kaposi's sarcoma in the United States, 1973-1995

The causes of multiple myeloma (MM) are obscure, but a laboratory association was recently reported between MM and human herpesvirus 8 (HHV-8), the probable etiologic agent of Kaposi's sarcoma (KS). Although there has been some additional laboratory corroboration, most laboratory studies have found no association between MM and HHV-8. We looked for indirect evidence of an HHV-8/MM association by evaluating whether MM is associated with KS in the United States. Cancer incidence and survival data were obtained from the Surveillance, Epidemiology, and End Results (SEER) program for the years 1973-1995. Strength of association was assessed for a number of cancer pairs using standardized incidence ratios (SIRs) (observed/expected double cancers). KS was strongly associated (SIR > 15) with non-Hodgkin's lymphoma and anal cancer, was modestly associated (2.5 < SIR < 5.5) with MM, Hodgkin's disease, and testicular cancer and was not significantly associated with 6 other cancers. Besides being associated with KS, MM was weakly associated (1.7 < SIR < 2.3) with Hodgkin's disease and testicular cancer. The SIRs for 7 other cancers paired with MM were all less than 1.6. Factors that might be responsible for the KS/MM association include MM-related immune dysfunction, HIV and HHV-8, but the role of these factors cannot be directly assessed through the SEER database. Although we cannot rule out the possibility that HHV-8 is linked to a small proportion of MM cases, the modest KS/MM association is evidence that the vast majority of MM cases are not likely to be associated with HHV-8.

Using case-control designs for genome-wide screening for associations between genetic markers and disease susceptibility loci

We used a case-control design to scan the genome for any associations between genetic markers and disease susceptibility loci using the first two replicates of the Mycenaean population from the GAW11 (Problem 2) data. Using a case-control approach, we constructed a series of 2-by-3 tables for each allele of every marker on all six chromosomes. Odds ratios (ORs) and 95% confidence intervals (95% CI) were estimated for all alleles of every marker. We selected the one allele for which the estimated OR had the minimum p-value to plot in the graph. Among these selected ORs, we calculated 95% CI for those that had a p-value < or = adjusted alpha level. Significantly high ORs were taken to indicate an association between a marker locus and a suspected disease-susceptibility gene. For the Mycenaean population, the case-control design identified allele number 1 of marker 24 on chromosome 1 to be associated with a disease susceptibility gene, OR = 2.10 (95% CI 1.66-2.62). Our approach failed to show any other significant association between case-control status and genetic markers. Stratified analysis on the environmental risk factor (E1) provided no further evidence of significant association other than allele 1 of marker 24 on chromosome 1. These data indicate the absence of linkage disequilibrium for markers flanking loci A, B, and C. Finally, we examined the effect of gene x environment (G x E) interaction for the identified allele. Our results provided no evidence of G x E interaction, but suggested that the environmental exposure alone was a risk factor for the disease.

Whole genome association studies for genes affecting alcohol dependence

We applied the transmission/disequilibrium test (TDT) for sibs (S-TDT) and for families with one parent (1-TDT), to the Collaborative Study on the Genetics of Alcoholism data set. The combined test is used to screen the whole genome to locate genes responsible for alcohol dependence. This analysis supports the previous finding that the region close to GABRB1 on chromosome 4 might be associated with alcohol dependence. The regions close to D6S474 and D11S1998 are also of particular interest. We found segregation distortion at the GR1K1 locus. The segregation distortion might be due to the binning method used in genotyping at this locus.

Can dialysis therapy be improved? A report from the ESRD Core Indicators Project

We assessed the association between quality improvement interventions conducted during the End-Stage Renal Disease (ESRD) Core Indicators Project and changes in the adequacy of hemodialysis between 1993 and 1996. Improvement of hemodialysis adequacy was measured by baseline and annual urea reduction ratios (URRs) in representative samples of ESRD Network patients. Random samples of in-center hemodialysis patients aged 18 years and older who had received hemodialysis during the fourth quarters of 1993, 1994, 1995, and 1996 were used to calculate Network-specific outcomes. A mean URR was calculated for each patient using the first pretreatment and posttreatment blood urea nitrogen for October, November, and December of each study year. Both national and Network-specific interventions were used to provide feedback reports and technical assistance to treatment centers to foster improvement in hemodialysis adequacy. All Networks distributed reports on the patterns of treatment center URR levels and physician and patient educational materials to each center in the Network. Each Network selected an annual 10% sample of treatment centers in 1994 and 1995 and conducted quality improvement activities to assist the selected centers to improve dialysis adequacy. We defined Network-specific interventions by a survey of the 18 Networks conducted during 1995 to determine the characteristics of Network-specific activities used to improve adequacy of hemodialysis. The outcome of interest was the change over time in Network-specific URR value. Sustained improvement in the URR occurred within all 18 Networks between 1993 and 1996. The mean national URR increased from 62.7% in 1993 to 66. 8% in 1996. The proportion of patients with URR >/= 65% increased from 43% in 1993 to 68% in 1996. Networks reported implementing a variety of intervention strategies that included educational activities, continuous quality improvement workshops, on-site assistance, and supervision of selected treatment facilities until care improved. Network-specific interventions independently associated with an increased rate of improvement in URR included prolonged supervision of the selected facilities. We concluded that the sustained improvement in hemodialysis care that occurred after the inception of the ESRD Core Indicators Project was associated with specific ESRD Network interventions.

Evaluating the exposure and disease relationship with adjustment for different types of exposure misclassification: a regression approach

Misclassification of exposure can lead to biased results in the epidemiologic research. Available methods accounting for misclassification often require the use of a gold standard or assume non-differential misclassification of exposure. We present a regression approach which can detect and account for different types of misclassification when estimating the exposure and disease relationship. This approach uses two imperfect measures of a dichotomous exposure and does not require a gold standard. Standard statistical packages with a logistic regression module can be used for estimation of parameters through the EM algorithm process. Two examples are used to illustrate the methodology.

Transmission disequilibrium test (TDT) when only one parent is available: the 1-TDT

The transmission disequilibrium test (TDT) is a useful method to locate mutations linked to disease genes associated with complex diseases. TDT requires genotypes of affected individuals and their parents. Recently, Ewens and Spielman (Am J Hum Genet 1998;62:450-8) extended the TDT for use in sibships with at least one affected and one unaffected individual and devised a new test called the sib transmission/disequilibrium test (S-TDT). The S-TDT can be applied to diseases with late age at onset such as non-insulin-dependent diabetes mellitus, psychiatric disorders, and diseases related to aging. For some disorders, it might be relatively easy to obtain the genotype of one parent either because the other parent is not available for study or he/she is not cooperative. Curtis and Sham (Ann Hum Genet 1995;59:323-36) showed that bias in transmitting certain alleles is introduced if only heterozygous parents and homozygous offspring are used in the TDT. In this paper, the authors propose a new test, the 1-TDT, to detect linkage between a candidate locus and a disease locus using genotypes of affected individuals and only one available parent for each affected individual. The test is not biased under the null hypothesis of no linkage or association. The authors validate their test using both simulated and real data sets. Finally, they show how to combine data from different types of families.

Effects of temperature and snowfall on mortality in Pennsylvania

The relation between exposure to severe cold weather and mortality is examined in a retrospective study of deaths occurring during the month of January from 1991 to 1996 in Pennsylvania. Using division-days as units of observation (n = 1,560) aggregated from death certificates and geographic divisions, the authors estimated mortality rates for total deaths and deaths due to ischemic heart disease, cerebrovascular diseases, and respiratory diseases by analyses based on generalized estimating equations. Total mortality increased on days of "extreme" climatic conditions, that is, when snowfall was greater than 3 cm and when temperatures were below -7 degrees C (rate ratio (RR) = 1.27, 95 percent confidence interval (CI) 1.12-1.44). On days of extreme conditions, mortality due to ischemic heart diseases tripled among males aged 35-49 years (RR = 3.54, 95 percent CI 2.35-5.35), increased for men aged 50-64 years (RR = 1.77, 95 percent CI 1.32-2.38), and rose for males aged 65 years and older (RR = 1.58, 95 percent CI 1.37-1.82), when compared with milder conditions. Among females, mortality for those aged 65 years and older increased for respiratory causes (RR = 1.68, 95 percent CI 1.28-2.21) and cerebrovascular causes (RR = 1.47, 95 percent CI 1.13-1.91). Cold and snow exposure may be hazardous among men as young as 35 years.

Validation of the pneumonia severity index. Importance of study-specific recalibration

OBJECTIVE

To evaluate the predictive validity and calibration of the pneumonia severity-of-illness index (PSI) in patients with community-acquired pneumonia (CAP).

PATIENTS

Randomly selected patients (n = 1,024) admitted with CAP to 22 community hospitals.

MEASUREMENTS AND MAIN RESULTS

Medical records were abstracted to obtain prognostic information used in the PSI. The discriminatory ability of the PSI to identify patients who died and the calibration of the PSI across deciles of risk were determined. The PSI discriminates well between patients with high risk of death and those with a lower risk. In contrast, calibration of the PSI was poor, and the PSI predicted about 2.4 times more deaths than actually occurred in our population of patients with CAP.

CONCLUSIONS

We found that the PSI had good discriminatory ability. The original PSI overestimated absolute risk of death in our population. We describe a simple approach to recalibration, which corrected the overestimation in our population. Recalibration may be needed when transporting this prediction rule across populations.

Excess hospital admissions during the July 1995 heat wave in Chicago

INTRODUCTION

This study describes medical conditions treated in all 47 non-VA hospitals in Cook County, IL during the 1995 heat wave. We characterize the underlying diseases of the susceptible population, with the goal of tailoring prevention efforts.

METHODS

Primary and secondary discharge diagnoses made during the heat wave and comparison periods were obtained from computerized inpatient hospital discharge data to determine reasons for hospitalization, and comorbid conditions, respectively.

RESULTS

During the week of the heat wave, there were 1072 (11%) more hospital admissions than average for comparison weeks and 838 (35%) more than expected among patients aged 65 years and older. The majority of this excess (59%) were treatments for dehydration, heat stroke, and heat exhaustion; with the exception of acute renal failure no other primary discharge diagnoses were significantly elevated. In contrast, analysis of comorbid conditions revealed 23% (p = 0.019) excess admissions of underlying cardiovascular diseases, 30% (p = 0.033) of diabetes, 52% (p = 0.011) of renal diseases, and 20% (p = 0.027) of nervous system disorders. Patient admissions for emphysema (p = 0.007) and epilepsy (p = 0.009) were also significantly elevated during the heat wave week.

CONCLUSIONS

The majority of excess hospital admissions were due to dehydration, heat stroke, and heat exhaustion, among people with underlying medical conditions. Short-term public health interventions to reduce heat-related morbidity should be directed toward these individuals to assure access to air conditioning and adequate fluid intake. Long-term prevention efforts should aim to improve the general health condition of people at risk through, among other things, regular physician-approved exercise.

Occupational risk factors for sarcoma subtypes

Herbicides, chlorophenols, and other occupational exposures are suspected risk factors for soft-tissue sarcoma, but the epidemiologic evidence is inconsistent. Given that soft-tissue sarcomas represent a heterogeneous mix of cancer subtypes and that these subtypes have different disease patterns by race, sex, and age at diagnosis, studying all soft-tissue sarcomas combined may mask subtype-specific associations. Using the Selected Cancers Study, a large population-based case-control study of sarcoma conducted among U.S. men aged 30 to 60 in 1984 to 1988, we explored the occupational risk factors for soft-tissue sarcoma subtypes and skeletal sarcoma. The analysis included 251 living sarcoma cases (48 dermatofibrosarcoma protuberans, 32 malignant fibrohistiocytic sarcoma, 67 leiomyosarcoma, 53 liposarcoma, and 51 skeletal sarcoma) and 1908 living controls. Exact conditional logistic regression models suggested patterns of subtype specificity for occupational exposures. Self-reported herbicide use was associated with malignant fibrohistiocytic sarcoma (OR = 2.9, 95% CI = 1.1-7.3). We found elevated risks for chlorophenol exposure and cutting oil exposure and malignant fibrohistiocytic sarcoma and leiomyosarcoma. We found no occupational risk factor for liposarcoma. Polytomous regression models identified different odds ratios across subtypes for plywood exposure and exposure to wood and saw dust. Although exploratory, this analysis suggests that occupational risk factors for sarcoma are not uniform across subtypes.

Case-only design to measure gene-gene interaction

The case-only design is an efficient and valid approach to screening for gene-environment interaction under the assumption of the independence between exposure and genotype in the population. In this paper, we show that the case-only design is also a valid and efficient approach to measuring gene-gene interaction under the assumption that the frequencies of genes are independent in the population. Just as the case-only design requires fewer cases than the case-control design to measure gene-environment interaction, it also requires fewer cases to measure gene-gene interactions.

A comparison of the Charlson comorbidity index derived from medical record data and administrative billing data

The objective of this article is to compare the Charlson comorbidity index derived from medical record data (Chart Index) with the same index derived from billing data (ICD-9 Index) to determine how well each predicted inpatient and 30-day mortality, length of stay, and complications among Medicare beneficiaries hospitalized for carotid endarterectomy. Economic and time constraints have increased the need for risk adjusters derived from administrative data, yet few studies have compared these measures with those derived from chart review. Using logistic regression, the Chart Index was found to be a significant predictor of inpatient mortality, 30-day mortality, length of stay, and complications, after controlling for age, gender, and neurologic and medical risk factors (P values = 0.004, 0.056, 0.0001, and 0.042, respectively). The ICD-9 Index approached significance as a predictor of the outcomes (P values = 0.092, 0.100, 0.093, and 0.080, respectively). The Chart Index was shown to be superior to the ICD-9 Index within this patient sample.

Mycobacterium sp. as a possible cause of hypersensitivity pneumonitis in machine workers

Hypersensitivity pneumonitis (HP) in workers exposed to metal removal fluids (MRFs) is increasing. This study supports the hypothesis that aerosolized mycobacteria colonizing the MRFs likely cause the disease. Three case studies of HP outbreaks among metal workers showed potentially high exposures to a rare and newly proposed Mycobacterium species. Retrospective review of samples submitted to our laboratory showed an association between presence of mycobacteria and HP.

A new method for estimating the risk ratio in studies using case-parental control design

The authors describe a new simple noniterative, yet efficient method to estimate the risk ratio in studies using case-parental control design. The new method is compared with two other noniterative methods, Khoury's method and Flanders and Khoury's method, and with a maximum likelihood-based method of Schaid and Sommer. The authors found that the variance of the new estimation method is usually smaller than that of Khoury's method or Flanders and Khoury's method and that it is slightly larger than that of the maximum likelihood-based method of Schaid and Sommer. Despite the slightly large variance of the new estimator compared with that of the maximum likelihood-based method, the simplicity of the new estimator and its variance makes the new method appealing. When genotypic information for only one parent is available, the authors also describe a method to estimate the risk ratio without assuming Hardy-Weinberg equilibrium or random mating. A simple formula for the variance of the estimator is given.

Mortality in end-stage renal disease is associated with facility-to-facility differences in adequacy of hemodialysis

Death rates of end-stage renal disease (ESRD) patients treated with hemodialysis vary substantially among treatment centers. The association between facility-to-facility differences in delivered hemodialysis dose and facility-specific mortality rates was examined among 5817 randomly selected patients treated with hemodialysis on October 1, 1994, from all 213 hemodialysis treatment centers reporting to ESRD Network 6. The mean urea reduction ratio (URR) for each treatment center, a measure of hemodialysis adequacy, was calculated for each facility, using measurements made by center staff members during one treatment for each of the randomly selected patients. During 7 mo of follow-up (ending April 30, 1995), 441 (7.6%) patients died. The average URR among the treatment centers was 64.9%. There was a strong, inverse association between increasing treatment center URR and adjusted mortality count (P = 0.009). Other treatment center characteristics associated with increased mortality included free-standing status (P = 0.009) and decreasing frequency of reported physician supervision of care (P = 0.01). It was concluded that lower average levels of dialysis adequacy in treatment centers are associated with higher rates of death, and this association persists after controlling for facility-to-facility differences in patient and nonpatient characteristics.

Family history score as a predictor of breast cancer mortality: prospective data from the Cancer Prevention Study II, United States, 1982-1991

A consistent predictor of a woman's risk for breast cancer is a family history of the disease. Most studies of family history and breast cancer have used the number of affected relatives in the family to calculate relative risk, but they have not considered the heterogeneity of the familial risk for breast cancer in a systematic way. With the use of data from a large prospective mortality study of US adults, the authors compared simple classification of family history of breast cancer (yes/no) to the method of using a quantitative family history score method, which takes into account the effects of family structure, age, and birth cohort as predictors of breast cancer mortality. After 9 years of follow-up, 1,428 cases of fatal breast cancer were observed among 453,073 women with complete information on number and age of siblings and family history. With the use of the family history score, about one-third of women with a positive family history of breast cancer were at no higher risk for breast cancer mortality than those without a family history of the disease. As a quantitative measure of relative risk for each family, family history score gave a better fit to the data, and it provided an incremental improvement of predictive accuracy of developing fatal breast cancer. Family history score can also be used as a categorical variable to stratify families. This allows researchers to focus on which risk groups would benefit from conducting further genetic analysis and to test the effects of genetic factors, environmental exposure, and gene-environment interactions on the etiology of the development of breast cancer.

Carotid endarterectomy among Medicare beneficiaries: a statewide evaluation of appropriateness and outcome

BACKGROUND AND PURPOSE

We sought to examine the appropriateness and the surgical outcomes of carotid endarterectomy (CE) in unselected community hospitals to identify opportunities for improvement.

METHODS

We performed a retrospective review of all CEs performed on Medicare beneficiaries in Georgia in 1993 (n = 1945). Conclusions regarding appropriateness were based on current guidelines as interpreted by a physician reviewer and were supported by the aggregate results of structured, blinded overreading by clinicians with relevant expertise. Adverse outcomes were confirmed and rated as to severity by a physician. Outcomes were correlated with demography, vascular anatomic findings, comorbidity, surgical techniques, and hospital characteristics.

RESULTS

The majority of the patients (51%) were asymptomatic at presentation. CEs were performed appropriately in 96.1% of the cases in accordance with current guidelines. There was no significant difference in the rate of appropriateness between the symptomatic (96%) and the asymptomatic patients (96.4%). Survival without stroke or myocardial infarction (MI) was 94.3%. The 30-day mortality was 1.9%; moderate to severe strokes occurred in 1.8%, stroke-related death in 0.7%, MI in 1.1%, and MI-related death in 0.5%. Those hospitals performing <10 CEs in the observed year had a statistically significant higher morbidity and mortality as well as an increase in less severe complications such as hematomas, wound dehiscence, wound infection, and pneumonia than did hospitals with higher volume of CEs.. Older patients and women had statistically significantly higher morbidity and mortality. Patients with a Charlson Severity Index score of > or =1 had a risk for adverse outcomes 3.4 times higher than patients with a score of 0 after adjustment for age and sex.

CONCLUSIONS

The great majority of CEs performed in Georgia on Medicare patients were appropriate, according to current guidelines. Slightly more than half of the patients were asymptomatic as defined in the Asymptomatic Carotid Atherosclerosis Study. In hospitals performing <CEs in the index year, we noted higher morbidity and mortality, as well as an increase in less severe complications. This relationship between the volume of surgery and outcome was confirmed in the analysis of the 30-day mortality for all Medicare cases (n = 10,569) performed in Georgia from 1991 to 1995. This was the most important opportunity for improvement identified in this study. In view of the increased use of CE, it is important to continue to monitor the patterns of practice for this procedure to improve outcomes.