Air pollution and human health: a review and reanalysis.

Since 1970, Lave and Seskin have published a series of articles dealing with the question, "Does air pollution shorten lives?" Their recent book reports revised and extended analyses of their previous studies emphasizing policy implications. We have undertaken a review of Lave and Seskin's book to evaluate the methodology used and hence gain some insight into the strength of the conclusions reached. This review concentrates on methodology and its application to establishing and quantifying the association between air quality and health. Beyond simply reviewing the analyses reported in Lave and Seskin's book, we have duplicated and expanded two of the reported analyses. Our detailed reanalysis is presented both to verify reported results, and to illustrate the difficulties encountered in such an analysis. Our overall conclusion is that Lave and Seskin have done a thorouth job of reporting and interpreting the various analyses that they performed. Lave and Seskin have made a pioneering effort in showing an association between mortality rates and air pollution. We do not disagree with the conclusion of the existence of an association but have some reservations about their methods of estimating its magnitude. We were particularly concerned that Lave and Seskin did not fully investigate how well their models fit these data. Our reanalysis results in estimated effects which differ considerably from the values reported by Lave and Seskin. Thus, we conclude that the regression coefficients are quite unstable and so must be used with care. Assessing the relative costs and benefits of reducing air pollution without extensive sensitivity analysis could, therefore, be misleading.

Plasma thromboxane and pulmonary artery pressure in neonates treated with extracorporeal membrane oxygenation.

To examine whether neonates with persistent pulmonary hypertension are subject to a thromboxane-mediated exacerbation of their pulmonary hypertension during extracorporeal membrane oxygenator therapy (a form of partial cardiopulmonary bypass), we performed serial measurements of plasma thromboxane B2 and pulmonary artery pressure before, during, and after extracorporeal membrane oxygenation. Pulmonary artery pressure was high before extracorporeal membrane oxygenation, did not increase after the start of this therapy, but began to decrease after 48 hours of extracorporeal membrane oxygenation. During the course of extracorporeal membrane oxygenation, mean pulmonary artery pressure decreased by 50% and mean plasma thromboxane B2 levels decreased by 70%. In addition, serial plasma thromboxane B2 levels were significantly correlated with pulmonary artery pressures in individual infants with a primary diagnosis of meconium aspiration (r = 0.965 to 0.723). We speculate that the decrease in pulmonary artery pressure and plasma thromboxane B2 levels over time may reflect resolution of acute lung injury and that thromboxane B2 may play a role in regulating pulmonary artery pressure in infants with meconium aspiration.

Rate of deterioration in accommodative esotropia correlated to the AC/A relationship.

We analyzed the claim that accommodative esotropia tends to deteriorate with greater frequency if the accommodation convergence relationship (AC/A) is high. Records of 119 patients whose eyes were aligned with spectacles alone were studied. Their AC/A relationships were graded according to the difference between the distance and near measurements: normal included 0 to 9 prism diopters (delta) difference; grade 1 ranged from 10 to 19 delta difference; grade 2 from 20 to 29 delta difference; and in grade 3 the difference was 30 delta or greater. Deterioration is characterized by a nonaccommodative component of esotropia greater than 10 delta at distance becoming superimposed on the initial accommodative esotropia. Deterioration occurred in 7.7% of patients with a normal AC/A, 25% with grade 1 high AC/A, 44% with grade 2 high AC/A, and 52% with grade 3 high AC/A. Hypotheses were investigated using chi square, t-test, analysis of variance (ANOVA), and log linear analyses. Distributional differences were highly significant by chi square test (p = 0.001) with a rejection of the null hypothesis of no difference between the groups at the alpha = 0.05 level. An alternate analysis of average AC/A ratio in the deteriorated versus nondeteriorated patients was equally statistically significant by the t-test. Hypermetropia was significantly higher in the normal AC/A group. Multi-factor comparisons showed that time-to-deterioration, treatment delay, age of onset, and amblyopia were factors that did not relate significantly to the incidence of deterioration.

Statistical procedures for assignments of upper bounds for radiation doses to an unbadged individual in a group for which some individuals are badged.

Consider a group of individuals exposed to radiation for an interval of time, such as units of military personnel during nuclear weapons tests three and four decades ago. Often, dosimetry information is known for several, but not all, individuals of that group. A nonparametric procedure is described that gives probability-based, upper-bound estimates of doses for an unbadged individual (within the group) for whom dose information is required; such information is useful in litigation. The method is illustrated by an application to a Navy unit.

Predictors of survival and characteristics of childhood stroke.

To determine predictors of outcome we reviewed 226 medical records of patients admitted to Children's National Medical Center with the diagnosis of cerebrovascular disease from 1978 to 1988. Ninety-five cases of stroke were identified by either neuroimaging techniques (87), autopsy (7), or clinical examination (1). Causal factors implicated in 89% of the patients included infectious (21%), vascular (18%), hematologic (15%), cardiac (13%) problems, minor trauma (8%) or miscellaneous (14%) causes. Patient outcome (n = 88) included residual impairment in 54%, complete resolution of their initial deficit in 23%, and death in 23%. Among patients with abnormal CT findings (n = 60), a logistic regression model revealed that patients with hemorrhage were at a significant risk (p = 0.0469) for death (odds ratio (OR) = 5.5, 95% confidence interval (CI) 1.2-24.5); those with an altered level of consciousness (stupor or coma) on presentation were also at risk (p = 0.0166; OR = 6.94, CI 1.7-28.5). The sensitivity and specificity of this model were 57% and 93%, respectively. No other clinical, laboratory, or demographic variable analyzed was predictive of outcome.

Survival effects of ZDV, ddI, and ddC in patients with CD4 < or = 50 cells/mm3.

Seven major clinical trials for the treatment of HIV-infected individuals are investigated. The treatments used in these trials were zidovudine, dideoxyinosine, dideoxycytosine, and one combination for patients with CD4 cell counts < 500 cells/mm3. Patients in each trial are partitioned into two subgroups based on their baseline CD4 cell counts: patients with CD4 < or = 50 cells/mm3 and patients with CD4 > 50 cells/mm3. The difference between treatment effects, using survival as a measure of effect, within each subgroup is calculated separately for each trial; this difference is referred to as "subgroup response." For each trial the difference between the subgroup responses is calculated and standardized. A meta-analysis is conducted over all seven trials for the differences between subgroup responses; the consistency of responses is evaluated across all trials among patients within baseline CD4 subgroups. Based on the result of this meta-analysis we conclude that there is no evidence that the treatment effects in patients with CD4 < or = 50 cells/mm3 are different from those among patients with CD4 > 50 cells/mm3. This result is observed in patients with different antiviral experience and different baseline characteristics. Risk ratios as well as chi 2 statistics are used to quantify the response rates in different subgroups. Kaplan-Meier curves of survival for these trials are depicted for all patients and each subgroup separately. In most of the trials the Kaplan-Meier curves for the patients with CD4 < or = 50 cells/mm3 resemble those for all patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Cerebrospinal fluid and plasma beta-endorphin-like immunoreactivity in full-term neonates and in preterm neonates with and without apnea of prematurity.

Beta-endorphin-like immunoreactivity (B-ELI) was measured in cerebrospinal fluid (CSF) and plasma from infants of postnatal age 24 h to 70 days. Three groups were examined: 17 were of postconceptual age greater than or equal to 37 weeks, 16 were postconceptual age less than or equal to 35 weeks without apnea and 10 were of postconceptual age less than or equal to 35 weeks with apnea. All infants were clinically stable. Two-way analysis of variance between groups showed no difference in the concentration of B-ELI in CSF or plasma, or in the CSF/plasma B-ELI ratio. Concentrations of B-ELI in plasma were significantly higher in infants of postnatal age 1-3 weeks and greater than or equal to 4 weeks, than in infants of postnatal age less than 1 week. We conclude that, in nonstressed infants, there is no relationship between the concentration of B-ELI in CSF or plasma and a concurrent diagnosis of apnea of prematurity. Our data indicate that a significant developmental increase occurs in the plasma concentration of B-ELI after the first week of life.

Assessing periodontal pathogens in children with varying levels of oral hygiene.

This study investigated the presence of organisms Actinobacillus actinomycetecomitans (A.a.), Bacteroides intermedius (B.i.), and Bacteroides gingivalis (B.g.), believed to be associated with periodontal disease etiology in children. The study included fifty children, aged six to fourteen years, selected from patients seeking routine dental care at Children's Hospital National Medical Center. A Modified Plaque Index (MPI) denoting the clinical presence and severity of the periodontal disease was obtained for each patient and ranged from 0-3. A total of 200 samples were obtained from the gingival crevice of the mesial surface of all first permanent molars and tested with DNA probes for the presence of three microorganisms. No correlations were found between race, sex, age, and the Modified Plaque Index. The results show that A. actinomycetecomitans was found only in patients with MPI greater than or equal to 1. B. gingivalis was present only in combination with another organism.

High Ia (HLA-DR) and low CD11b (Mo1) expression may predict early conversion to leukemia in myelodysplastic syndromes.

The FAB classification of myelodysplastic syndromes (MDS) has been useful in predicting prognosis; however, additional methods are required to detect patients at high risk for early conversion to acute nonlymphoblastic leukemia (ANLL). Using a panel of monoclonal antibodies to myelomonocytic surface antigens (MMSA) and flow cytometry, we studied bone marrow cells from 26 patients with MDS of all five FAB subtypes. The MMSA studied included Ia (HLA-DR), CD11b (Mo1), CD14 (Mo2, My4), CD13 (My7), and CD33 (My9). Marrows were considered "positive" for a given MMSA if the percentage of reactive cells exceeded the upper limit of the normal range. Twenty-four of twenty-six patients (92.3%) were CD13 (My7)+, suggesting that CD13 may serve as a diagnostic marker for MDS. Ten of twelve patients who developed ANLL during a median follow-up of 44 weeks were Ia(HLA-DR)+. The Kaplan-Meier estimated median time to leukemia (TTL) was 16 weeks for Ia+ patients and 88 weeks for Ia- patients (P = 0.004). All six patients who developed ANLL before 16 weeks from diagnosis were Ia+, while none of the Ia- patients converted to ANLL before 24 weeks. Nine of thirteen patients with low CD11b (Mo1) expression (< 53% reactive cells) developed ANLL, compared with only two of 11 patients with high CD11b expression (> 53% reactive cells). Kaplan-Meier estimated TTL was 29 weeks for patients with low CD11b, compared to 160 weeks for patients with high CD11b (P < 0.05). Patients who met both criteria, Ia+ and low CD11b, represented the poorest prognostic subgroup, with median TTL of 13 weeks compared with 88 weeks for the others (P = 0.017). Ia and CD11b patterns were not specific for MDS subtype, and their expression did not correlate with blast count. These data suggest that MDS patients whose bone marrow cells demonstrate high Ia (HLA-DR) and low CD11b (Mo1) expression represent a poor prognostic subgroup with short TTL. These patients may be candidates for early aggressive or investigational treatment.

Surfactant protein A concentrations in tracheal aspirate fluid from infants requiring extracorporeal membrane oxygenation.

To understand the lung abnormalities leading to respiratory failure in infants, we measured 35,000-dalton surfactant protein A concentrations in tracheal aspirate fluid collected daily from 25 infants receiving extracorporeal membrane oxygenation (ECMO). Surfactant protein A concentrations were standardized per milligrams of total protein present in the aspirate. Among the 23 survivors with complete data, the surfactant protein A concentration increased significantly with time (p less than 0.0001). Concurrent increases in lung compliance (p less than 0.0001) and radiographic scores (p less than 0.0001) were also observed. This increase in surfactant protein A content may reflect lung recovery from barotrauma and oxygen toxic effects or be a response to the primary pulmonary disease process. The two infants who did not survive extracorporeal membrane oxygenation failed to demonstrate these trends.