Predicting asthma severity from allergic sensitivity to cockroaches in pregnant inner city women.

OBJECTIVE: To measure and compare cockroach (CR)-specific immunoglobin E (IgE) in sera from pregnant women with mild, moderate and severe asthma. STUDY DESIGN: CR IgE levels were measured in stored sera collected during the Collaborative Perinatal Project. Three matched groups of 93 women were formed: group I (mild), history of asthma but no acute exacerbation; group II (moderate), acute asthma exacerbation; group III (severe), required hospitalization for a diagnosis of status asthmaticus. ANOVA was used to compare the three means. RESULTS: Mean CR IgE paralleled prenatal asthma severity. Mean values were 6.50, 13.12 and 28.99 kU/L for groups I, II and III, respectively (P = .06). High allergen sensitivity, defined as CR IgE > 60 kU/L, was identified in 8 of the 93 study samples. The prevalence of high allergen sensitivity increased as clinical asthma became more severe. Sixty-two percent (5/8) of the high allergen sensitivity occurred in group III. CONCLUSION: There appears to be a positive correlation between sensitivity to CR allergens and asthma severity during pregnancy, and these findings support further evaluation of CR allergen sensitivity as a predictor of asthma severity in pregnancy.

Corticosteroid treatment for idiopathic facial nerve paralysis: a meta-analysis.

OBJECTIVE: A meta-analysis was designed to evaluate facial recovery in patients with complete idiopathic facial nerve paralysis (IFNP) by comparing outcomes of those treated with corticosteroid therapy with outcomes of those treated with placebo or no treatment. STUDY DESIGN: Meta-analysis of prospective trials evaluating corticosteroid therapy for idiopathic facial nerve paralysis. METHODS: A protocol was followed outlining methods for trial selection, data extraction, and statistical analysis. A MEDLINE search of the English language literature was performed to identify clinical trials evaluating steroid treatment of IFNP. Three independent observers used an eight-point analysis to determine inclusion criteria. Data analysis was limited to individuals with clinically complete IFNP. The endpoints measured were clinically complete or incomplete facial motor recovery. Effect magnitude and significance were evaluated by calculating the rate difference and Fisher's Exact Test P value. Pooled analysis was performed with a random effects model. RESULTS: Forty-seven trials were identified. Of those, 27 were prospective and 20 retrospective. Three prospective trials met the inclusion criteria Tests of heterogeneity indicate the trial with the smallest sample size (RD = -0.19; 95% CI, -0.58-0.20), to be an outlier. It was excluded from the final analysis. Analyses of data from the remaining two studies indicate corticosteroid treatment improves complete facial motor recovery for individuals with complete IFNP. Rate difference demonstrates a 17% (990% CI, 0.01-0.32) improvement in clinically complete recovery for the treatment group based on the random effects model. CONCLUSIONS: Corticosteroid treatment provides a clinically and statistically significant improvement in recovery of function in complete IFNP.

Maternal serum paraxanthine, a caffeine metabolite, and the risk of spontaneous abortion.

BACKGROUND: Whether the consumption of caffeine during pregnancy increases the risk of spontaneous abortion is controversial. Prior studies have determined caffeine consumption by questionnaire. We used a biologic marker, such as serum paraxanthine, a metabolite of caffeine, to measure the dose of caffeine. METHODS: In a nested case-control study, we measured serum paraxanthine in 591 women who had spontaneous abortions at less than 140 days' gestation and in 2558 matched women from the same clinic who gave birth to live infants at 28 weeks' gestation or later and who had serum drawn on the same day of gestation as the women who had abortions. The women were enrolled in the Collaborative Perinatal Project during the period from 1959 to 1966, and serum paraxanthine was measured over 30 years later. RESULTS: A total of 487 women who had spontaneous abortions (82 percent) and 2087 controls (82 percent) had quantifiable serum paraxanthine concentrations. However, the mean serum paraxanthine concentration was higher in the women who had spontaneous abortions than in the controls (752 vs. 583 ng per milliliter, P<0.001). The odds ratio for spontaneous abortion was not significantly elevated in the women who had serum paraxanthine concentrations of 1845 ng per milliliter or lower, corresponding to the 95th percentile of the matched women. However, the adjusted odds ratio for spontaneous abortion among women with serum paraxanthine concentrations higher than 1845 ng per milliliter, as compared with women who had concentrations below 50 ng per milliliter, was 1.9 (95 percent confidence interval, 1.2 to 2.8). CONCLUSIONS: Only extremely high serum paraxanthine concentrations are associated with spontaneous abortion. This suggests that moderate consumption of caffeine is unlikely to increase the risk of spontaneous abortion.

Resolving discrepancies between a meta-analysis and a subsequent large controlled trial.

CONTEXT: A recent meta-analysis found calcium supplementation to be highly effective in preventing preeclampsia but a large National Institutes of Health trial (Calcium for Preeclampsia Prevention [CPEP]) found no risk reduction due to calcium in healthy nulliparous women. OBJECTIVES: To resolve discrepancies between the results of the meta-analysis and the CPEP trial and to assess the role of effect heterogeneity in the discrepancies. DATA SOURCES: Literature search of English-language articles published prior to July 10, 1997, the date of publication of the CPEP trial, using MEDLINE and by a manual search of bibliographies of published articles. STUDY SELECTION: Trials were included if they reported data on preeclampsia and calcium supplementation. Fourteen trials were systematically evaluated for differences in study design and patient populations. One trial was excluded because its results were reported after publication of the major CPEP results. DATA EXTRACTION: The sample size and number of subjects who developed preeclampsia in the calcium supplementation group vs a control group were recorded and analyzed on an intent-to-treat basis. Each author independently extracted the data. DATA SYNTHESIS: Substantial heterogeneity existed across trials (P = .001). After stratifying studies by the presence of a placebo-controlled group and by high-risk and low-risk populations, the conclusions of the meta-analysis of placebo-controlled trials enrolling a low-risk population (relative risk, 0.79; 99% confidence interval, 0.44-1.42; P = .30) were compatible with the conclusions of the CPEP trial that calcium supplementation does not prevent preeclampsia in healthy nulliparous women. In contrast, the data implied a strong beneficial calcium effect (relative risk, 0.19; 99% confidence interval, 0.08-0.46; P = .001) in healthy high-risk subject populations. However, only 225 women were analyzed and because of inconsistent data, these results remain equivocal. CONCLUSIONS: Further studies are needed to establish the efficacy of calcium for preeclampsia prevention in healthy high-risk populations. A single summary measure does not adequately describe the findings of a meta-analysis when the observed effects in individual studies differ substantially. In such settings the primary focus should be to identify and incorporate pertinent covariates that reduce heterogeneity and allow for optimum treatment strategies.

Prostacyclin and thromboxane changes predating clinical onset of preeclampsia: a multicenter prospective study.

CONTEXT: An imbalance in vasodilating (prostacyclin [PGI2]) and vasoconstricting (thromboxane A2 [TxA2]) eicosanoids may be important in preeclampsia, but prospective data from large studies needed to resolve this issue are lacking. Because most trials using aspirin to reduce TxA2 production have failed to prevent preeclampsia, it is critical to determine whether eicosanoid changes occur before the onset of clinical disease or are secondary to clinical manifestations of preeclampsia. OBJECTIVE: To determine whether PGI2 or TxA2 changes occur before onset of clinical signs of preeclampsia. DESIGN, SETTING, AND PARTICIPANTS: Multicenter prospective study from 1992 to 1995 of subjects from the placebo arm of the Calcium for Preeclampsia Prevention Trial. Women who developed preeclampsia (n = 134) were compared with matched normotensive control women (n = 139). MAIN OUTCOME MEASURES: Excretion of urinary metabolites of PGI2 (PGI-M) and TxA2 (Tx-M) as measured from timed urine collections obtained prospectively before 22 weeks', between 26 and 29 weeks', and at 36 weeks' gestation. RESULTS: Women who developed preeclampsia had significantly lower PGI-M levels throughout pregnancy, even at 13 to 16 weeks' gestation (long before the onset of clinical disease); their gestational age-adjusted levels were 17% lower than those of controls (95% confidence interval [CI], 6%-27%; P=.005). The Tx-M levels of preeclamptic women were not significantly higher overall (9% higher than those of controls; 95% CI, -3% to 23%; P=.14). The ratio of Tx-M to PGI-M, used to express relative vasoconstricting vs vasodilating effects, was 24% higher (95% CI, 6%-45%) in preeclamptic women throughout pregnancy (P=.007). CONCLUSIONS: Our results show that reduced PGI2 production, but not increased TxA2 production, occurs many months before clinical onset of preeclampsia. Aspirin trials may have failed because an increase in thromboxane production is not the initial anomaly. Future interventions should make correcting prostacyclin deficiency a major part of the strategy to balance the abnormal vasoconstrictor-vasodilator ratio present in preeclampsia.

Deaths attributable to injuries in infants, United States, 1983-1991.

OBJECTIVE: To describe risk factors for injury death among infants in the United States by the specific external cause of death. METHODS: Data were analyzed from the US-linked birth/infant death files for the years 1983-1991. Potential risk factors for injury death were identified from birth certificate data and included both maternal and infant factors. Injury rates were calculated by external cause of death. Characteristics of infants who died from an injury were compared with those of the entire birth cohort. The independent effect of potential risk factors was assessed in multivariate analyses using a case-control study design. RESULTS: A total of 10 370 injury deaths were identified over the 9-year study period (29. 72/100 000 live births). The leading causes of death were homicide, suffocation, motor vehicle crashes, and choking (inhalation of food or objects). There was no significant temporal trend in the overall rate of injury death; however, this was because significant increases in the rates of death from homicide (6.4%/year) and mechanical suffocation (3.7%/year) were offset by decreases in rates of death from fires (-4.7%/year) and choking (-4.6%/year). In adjusted analyses, infants born to mothers with no prenatal care, <12 years of education, two or more previous live births, Native American race, or <20 years of age were at twice the risk of injury death compared with the lowest risk groups (initiation of prenatal care in the first trimester, >/=16 years of education, no previous live births, white, or >/=25 years of age). When analyzed by the specific cause of death, the factors that were associated most strongly with death varied. For example, Native Americans were at greatest risk of a motor vehicle related death (compared with whites: OR: 3.6; 95% CI: 1.8-7.1), and infants with birth weights of <1500 g were at greatest risk of death attributable to inhalation of food (compared with >/=2500 g: OR: 9.6; 95% CI: 3.3-28.0) or objects (OR: 11.8; 95% CI: 4.5-30.5). CONCLUSION: A number of sociodemographic characteristics are associated with an increased risk of injury-related death in infants. The strength of associations between specific risk factors and death varies with the external cause of death, thus identifying high-risk subgroups for targeting of cause-specific interventions and simultaneously increasing our understanding of the individual and societal mechanisms underlying these tragedies.

Epidural analgesia in association with duration of labor and mode of delivery: a quantitative review.

OBJECTIVE: This study was undertaken to quantitatively summarize previous literature on the effects of epidural analgesia in labor on the duration of labor and mode of delivery. STUDY DESIGN: Original studies published in English from 1965 through December 1997 were reviewed and assigned a quality score independently by 2 of the authors. Studies that met the minimal requirements were evaluated further. Data syntheses were performed separately according to study design and outcome measurements, including cesarean delivery, instrumental delivery, oxytocin augmentation, and durations of the first and second stages of labor. RESULTS: Seven randomized clinical trials and 5 observational studies met the minimal requirements. Among them 4 studies of each sort were included in the data synthesis. Both types of studies showed that epidural analgesia increased risk of oxytocin augmentation 2-fold. Clinical trials suggested that epidural analgesia did not increase the risk of cesarean delivery either overall or for dystocia, nor did it significantly increase the risk of instrumental vaginal delivery; however, observational studies reported a more than 4-fold increased risk of cesarean and instrumental deliveries. Although most studies showed a longer labor among women with epidural analgesia than without it, especially during the second stage, most of the studies used inappropriate statistical analysis. CONCLUSION: Epidural analgesia with low-dose bupivacaine may increase the risk of oxytocin augmentation but not that of cesarean delivery.

Serum cotinine concentration and self-reported smoking during pregnancy.

Although during pregnancy there is a better correlation between maternal serum cotinine concentration and adverse outcome than between self-reported smoking and such an outcome, few studies of pregnancy have measured cotinine concentration to determine how much a woman smokes. This study assessed the accuracy of self-reported smoking during pregnancy by performing serum cotinine assays on 448 women registered in the Collaborative Perinatal Project (1959-1966). Based on the assumption that a serum cotinine concentration of >10 ng/ml represented active smoking, 94.9% of women who denied smoking and 87.0% of women who stated that they smoked (kappa=0.83) reported their status accurately. Among smokers, the correlation coefficient between cotinine concentration and number of cigarettes smoked per day was 0.44. Serum cotinine concentration correlated more strongly than self-reported smoking with infant birth weight (r=0.246 vs. 0.200). In conclusion, this study showed that pregnant women accurately reported whether they smoked, but cotinine concentration was a better measure than self-report of the actual tobacco dose received.

Serum caffeine and paraxanthine as markers for reported caffeine intake in pregnancy.

PURPOSE: Previous studies of maternal caffeine use and pregnancy outcome have relied on self-reported use. Even if these were perfectly accurate, inter-individual differences in caffeine metabolism result in a relatively weak correlation between caffeine intake and serum concentration. The purpose of this study was to determine whether the serum concentration of caffeine or its primary metabolite, paraxanthine, obtained at an unknown time during working hours, is useful to distinguish between pregnant women who report consuming small and large amounts of caffeine. METHODS: We selected from the Birmingham fetal growth study 60 women with normal pregnancy outcomes who reported consuming < or = 0.8 mg/kg/day of caffeine in a 24-hour dietary recall, 60 who consumed 0.81-2.5 mg/kg/day, 60 who consumed 2.51-5.0 mg/kg/day and 59 who consumed > or = 5.01 mg/kg/day. These women had serum drawn for storage during regular clinic hours on the same day as the recall interview. Caffeine and paraxanthine were measured in the stored serum using high performance liquid chromatography. RESULTS: The weighted kappa coefficient between strata of caffeine intake and quartiles of serum paraxanthine was 0.58 among smokers and 0.53 among nonsmokers, versus 0.44 and 0.51, respectively, for quartiles of serum caffeine. The Pearson correlation coefficient between intake and paraxanthine was 0.50 for smokers and 0.53 for nonsmokers, and 0.37 and 0.51, respectively, for serum caffeine. These values are comparable to the correlation between reported smoking and serum cotinine in pregnancy. CONCLUSIONS: The serum concentrations of paraxanthine, and to a lesser degree, caffeine are useful to distinguish between women with varying levels of caffeine intake.

Trial of calcium to prevent preeclampsia.

BACKGROUND: Previous trials have suggested that calcium supplementation during pregnancy may reduce the risk of preeclampsia. However, differences in study design and a low dietary calcium intake in the populations studied limit acceptance of the data. METHODS: We randomly assigned 4589 healthy nulliparous women who were 13 to 21 weeks pregnant to receive daily treatment with either 2 g of elemental calcium or placebo for the remainder of their pregnancies. Surveillance for preeclampsia was conducted by personnel unaware of treatment-group assignments, using standardized measurements of blood pressure and urinary protein excretion at uniformly scheduled prenatal visits, protocols for monitoring these measurements during the hospitalization for delivery, and reviews of medical records of unscheduled outpatient visits and all hospitalizations. RESULTS: Calcium supplementation did not significantly reduce the incidence or severity of preeclampsia or delay its onset. Preeclampsia occurred in 158 of the 2295 women in the calcium group (6.9 percent) and 168 of the 2294 women in the placebo group (7.3 percent) (relative risk, 0.94; 95 percent confidence interval, 0.76 to 1.16). There were no significant differences between the two groups in the prevalence of pregnancy-associated hypertension without preeclampsia (15.3 percent vs. 17.3 percent) or of all hypertensive disorders (22.2 percent vs. 24.6 percent). The mean systolic and diastolic blood pressures during pregnancy were similar in both groups. Calcium did not reduce the numbers of preterm deliveries, small-for-gestational-age births, or fetal and neonatal deaths; nor did it increase urolithiasis during pregnancy. CONCLUSIONS: Calcium supplementation during pregnancy did not prevent preeclampsia, pregnancy-associated hypertension, or adverse perinatal outcomes in healthy nulliparous women.

Trial of Calcium for Preeclampsia Prevention (CPEP): rationale, design, and methods.

The results of ten clinical trials suggest that supplemental calcium may prevent preeclampsia. However, methodologic problems and differences in study design limit the acceptance of the results and their relevance to other patient populations. Many of the trials were conducted in countries where, unlike the United States, the usual daily diet contained little calcium. Moreover, none of the trials has reported the outcome of systematic surveillance for urolithiasis, a potential complication of calcium supplementation. In response to the need for a thorough evaluation of the effects of calcium supplementation for the prevention of preeclampsia in the United States, the trial of Calcium for Preeclampsia Prevention (CPEP) was undertaken at five university medical centers. Healthy nulliparous patients were randomly assigned to receive either 2 g supplemental calcium daily (n = 2295) or placebo (n = 2294) in a double-blind study. Study tablets were administered beginning from 13 to 21 completed weeks of gestation and continued until the termination of pregnancy. CPEP employed detailed diagnostic criteria, standardized techniques of measurement, and systematic surveillance for the major study endpoints and for urolithiasis. The nutrient intake of each patient was assessed at randomization and at 32-33 weeks gestation. This report describes the study rationale, design, and methods.

Meta-analysis in the design and monitoring of clinical trials.

Results from external studies often play an important role in many aspects of a clinical trial. Their incorporation into the decision making process of a trial, however, is rarely conducted in a formal manner. This conference will address what formal role, if any, meta-analytic summaries of external results should have in the design and monitoring of an ongoing trial. This introductory presentation describes in detail the example from obstetric research which motivated the conference topic, and, in a Bayesian framework, summarizes the general implications of formally incorporating meta-analytic results into the design and analysis of a new trial.

Problems in interpreting HIV sentinel seroprevalence studies.

Estimating human immunodeficiency virus (HIV) prevalence from sentinel seroprevalence studies is difficult. We characterize these studies and show that most are investigations of incompletely defined (hypothetical) cohorts and are usually based on nonprobability samples. Prevalence in HIV sentinel serosurveys is also time-averaged and vulnerable to several time-dependent sources of bias (e.g., migration, deaths, and changes in incidence). Assumptions must be made that these time-dependent biases did not meaningfully affect the data, and this can be helped by reducing the period of investigation. Furthermore, we show that "reliability" can not be adequately measured by standard error, that "internal validity" is vulnerable to self-selection bias and laboratory problems, and that "generalizability" is limited. We propose that what is needed is a procedure (like formal metaanalysis methods) incorporating information from several separate HIV sentinel seroprevalence studies, in a manner that is reproducible and can take into consideration the differences between studies.

Vasectomy and prostate cancer risk: methodological review of the evidence.

Two recent studies have reported a significantly elevated risk of prostate cancer among vasectomized men. To assess whether the new results conflict with earlier studies that found no significant overall association, and, if so, whether such a conflict could have a methodological basis, we reviewed the six major epidemiological studies of this topic. Statistical analysis revealed significant (p < 0.01) heterogeneity among the associations in the six studies, attributable to one of the recent studies. Scrutiny of the studies for fulfillment of eight methodological standards for scientific validity revealed that no study completely fulfilled more than four standards, and that all studies were deficient in avoiding detection bias and obtaining accurate vasectomy histories. Our review indicates that the evidence on this topic is indeed conflicting, that the quality of the evidence does not resolve the conflict, and that future studies of this topic, designed to ensure scientific credibility of results, are needed.

PIP: 2 recent studies have reported a significantly elevated risk of prostate cancer among vasectomized men. 6 major relevant epidemiological studies were reviewed to determine the new results conflicted with earlier studies finding no significant association and whether such conflict was attributable to methodological deficiencies. Statistical analysis indicated significant heterogeneity among the associations in the 6 studies (p 0.01) imputed to one of the studies. Examination of 8 methodological standards for scientific validity showed that none of the studies fulfilled more than 4 standards, and all were deficient in avoiding detection bias and obtaining accurate vasectomy histories. The evidence is conflicting concerning prostate cancer, and future studies are needed to ensure scientific credibility of results. The literature search included the MEDLINE bibliographic database from January 1970 to December 1991, which yield 6 studies on the link of vasectomy and risk of prostate cancer. The 8 criteria were prior hypothesis, single underlying suitable validated histories of vasectomy, protection against detection bias, valid diagnoses of prostate cancer, and incident prostate cancer outcomes. 1 was a cohort and 5 were case-control investigations during 1982-88. The follow-up after vasectomy varied: 15+, 20+, 30+, and 44 years with 40-86 years of documented ages. Analysis of homogeneity indicated similarity for 5 studies. Estimation of overall effect by odds ration estimates and confidence intervals suggested a slight risk of prostate cancer associated with vasectomy. Control for confounding by age was done in 5 studies, and no study was given credit for adequate protection against detection bias. 3 studies reported histological confirmation of prostate cancer, but none reported a review of diagnostic evidence validated with vasectomy histories.

An analysis of interobserver recognition of the histopathologic features of dysplastic nevi from a mixed group of nevomelanocytic lesions.

BACKGROUND: The diagnostic criteria for the histopathologic recognition of the dysplastic nevus are defined, but difficult to apply in practice. However, the use of skin biopsy is on the increase to search for this nevus because of a possible role in melanomagenesis. OBJECTIVE: The aim was to determine the degree of observer reproducibility in the detection of histologic nevomelanocytic dysplasia as measured across multiple observers and to test whether each observer was precise over a wide sample population. A second aim was to gain some measure of the threshold a person has for such lesions. METHODS: Histologic specimens from 50 coded nevomelanocytic tumors, which included a group of dysplastic nevi, were independently read by five observers in a blinded fashion. At study completion, the same 50 cases were read again by a melanoma reference pathologist and results analyzed by Cohen's kappa to assess agreement and interobserver variation. RESULTS: Corrected rates of agreement for dysplastic nevi ranged from 0.32 to 0.71. CONCLUSION: Continued study and experience are required for precision and reproducibility in the histologic recognition of dysplastic nevi.

Interpretation of linear regression models that include transformations or interaction terms.

In linear regression analyses, we must often transform the dependent variable to meet the statistical assumptions of normality, variance stability, or linearity. Transformations, however, can complicate the interpretation of results because they change the scale on which the dependent variable is measured. In this setting, the inclusion of product terms or the transformation of some independent (or predictor) variables may further complicate interpretation. In this article, we present some interpretations of linear models that include transformations or product terms. We illustrate these interpretations using regression analyses designed to study determinants of serum testosterone levels. These examples show how one can present results using simple measures, such as medians, and interpret regression parameters.

Closed-form estimates for missing counts in two-way contingency tables.

One method for analyzing contingency tables with missing observations is to model the missing-data mechanism using log-linear models. Previous methods for obtaining estimates (of missing counts and parameters) have required an iterative algorithm. In many cases, however, one can obtain estimates by use of a simple algebraic formula. We illustrate the method with data on smoking and birth weight.