General regression model for the subdistribution of a competing risk under left-truncation and right-censoring.

Left-truncation poses extra challenges for the analysis of complex time-to-event data. We propose a general semiparametric regression model for left-truncated and right-censored competing risks data that is based on a novel weighted conditional likelihood function. Targeting the subdistribution hazard, our parameter estimates are directly interpretable with regard to the cumulative incidence function. We compare different weights from recent literature and develop a heuristic interpretation from a cure model perspective that is based on pseudo risk sets. Our approach accommodates external time-dependent covariate effects on the subdistribution hazard. We establish consistency and asymptotic normality of the estimators and propose a sandwich estimator of the variance. In comprehensive simulation studies we demonstrate solid performance of the proposed method. Comparing the sandwich estimator with the inverse Fisher information matrix, we observe a bias for the inverse Fisher information matrix and diminished coverage probabilities in settings with a higher percentage of left-truncation. To illustrate the practical utility of the proposed method, we study its application to a large HIV vaccine efficacy trial dataset.

Penalized Q-Learning for Dynamic Treatment Regimens.

A dynamic treatment regimen incorporates both accrued information and long-term effects of treatment from specially designed clinical trials. As these trials become more and more popular in conjunction with longitudinal data from clinical studies, the development of statistical inference for optimal dynamic treatment regimens is a high priority. In this paper, we propose a new machine learning framework called penalized Q-learning, under which valid statistical inference is established. We also propose a new statistical procedure: individual selection and corresponding methods for incorporating individual selection within penalized Q-learning. Extensive numerical studies are presented which compare the proposed methods with existing methods, under a variety of scenarios, and demonstrate that the proposed approach is both inferentially and computationally superior. It is illustrated with a depression clinical trial study.

Doubly Robust Learning for Estimating Individualized Treatment with Censored Data.

Individualized treatment rules recommend treatments based on individual patient characteristics in order to maximize clinical benefit. When the clinical outcome of interest is survival time, estimation is often complicated by censoring. We develop nonparametric methods for estimating an optimal individualized treatment rule in the presence of censored data. To adjust for censoring, we propose a doubly robust estimator which requires correct specification of either the censoring model or survival model, but not both; the method is shown to be Fisher consistent when either model is correct. Furthermore, we establish the convergence rate of the expected survival under the estimated optimal individualized treatment rule to the expected survival under the optimal individualized treatment rule. We illustrate the proposed methods using simulation study and data from a Phase III clinical trial on non-small cell lung cancer.

An Exponential Bound for Cox Regression.

We present an asymptotic exponential bound for the deviation of the survival function estimator of the Cox model. We show that the bound holds even when the proportional hazards assumption does not hold.

Early immune response to the components of the type III system of Pseudomonas aeruginosa in children with cystic fibrosis.

The lungs of patients with cystic fibrosis (CF) are colonized initially by Pseudomonas aeruginosa, which is associated with progressive lung destruction and increased mortality. The pathogenicity of P. aeruginosa is caused by a number of virulence factors, including exotoxin A (ETA) and the type III cytotoxins (ExoS, ExoT, ExoU, and ExoY). P. aeruginosa contacts the plasma membrane to deliver type III cytotoxins through a channel formed by PopB, PopD, and PcrV; ETA enters mammalian cells via receptor-mediated endocytosis. The Wisconsin CF Neonatal Screening Project is a longitudinal investigation to assess the potential benefits and risks of newborn screening for CF; the project was the source of serum samples used in this study. Past studies evaluated the longitudinal appearance of antibodies to ETA and elastase and P. aeruginosa infections in patients with CF. The current study characterized the longitudinal appearance of antibodies to components of the type III system in children with CF. Western blot analyses showed that serum antibodies to PopB, PcrV, and ExoS were common. Longitudinal enzyme-linked immunosorbent assays determined that the first detection of antibodies to pooled ExoS/PopB occurred at a time similar to those of detection of antibodies to a P. aeruginosa cell lysate and the identification of oropharyngeal cultures positive for P. aeruginosa. This indicates that children with CF are colonized early with P. aeruginosa expressing the type III system, implicating it in early pathogenesis, and implies that surveillance of clinical symptoms, oropharyngeal cultures, and seroconversion to type III antigens may facilitate early detection of P. aeruginosa infections.

Acceleration of lung disease in children with cystic fibrosis after Pseudomonas aeruginosa acquisition.

As part of the ongoing Wisconsin Cystic Fibrosis (CF) Neonatal Screening Project, we had the unique opportunity to study the longitudinal relationship between Pseudomonas aeruginosa (Pa) acquisition and infection and developing lung disease in children with CF. The primary objective was to determine whether acquisition of Pa was associated with a measurable change in the progression of lung disease. Two outcome measures were used to study 56 patients who were diagnosed through newborn screening: 1) Wisconsin additive chest radiograph score (WCXR), based on the average of scores from a pulmonologist and a radiologist, and 2) the highest forced expired volume in 1 sec (FEV(1))/forced vital capacity (FVC) ratio. We used two measures of Pa acquisition: 1) time of first positive protocol-determined oropharyngeal (with cough) culture, and 2) the magnitude of antibody titer detected by ELISA assays, using as antigen a crude cell lysate, purified exotoxin A, or an elastase toxoid prepared from three Pa strains. Other predictor variables included age, pancreatic status, height-for age, and weight-for-age-percentiles. The best regression model for predicting changes in the WCXR included time to first positive culture and antibody titer for Pa elastase. Prior to Pa acquisition, WCXR worsened by 0.45 points/year (P > 0.25); after Pa acquisition, the rate of worsening increased significantly (P < 0.001) to 1.40 points/year. Each antibody titer level (log base 2) increased the score by 0.48 points (P < 0.001). The best regression model for predicting change in the FEV(1)/FVC included only time to first positive culture. Prior to Pa acquisition, the FEV(1)/FVC ratio declined by 1.29%/year; after Pa infection, the rate of decrease significantly accelerated to 1.81%/year (P = 0.001). Our data show that Pa acquisition is associated with declining pulmonary status in children with CF, and that this effect is probably gradual rather than precipitous. Because these patients were diagnosed and treated aggressively, our estimates of the effects of Pa acquisition may be conservative. We also conclude that the WCXR appears to be more sensitive than FEV(1)/FVC in detecting early changes in lung disease associated with CF.

Comparison of caloric intake and weight outcomes of an ad lib feeding regimen for preterm infants in two nurseries.

BACKGROUND: Effects on caloric intake and weight gain of an ad libitum (ad lib) feeding regimen for preterm infants may be specific to a special care nursery. OBJECTIVE: To explore across two nurseries the similarity of effect on caloric intake and weight gain of an ad lib feeding regimen compared with a prescribed regimen and the similarity of effect of caloric intake on weight gain. METHODS: All infants participating in the multi-site randomized clinical trial (RCT) of the ad lib feeding regimen were <35 weeks gestational age at birth and had birth weight appropriate for gestational age. Data on caloric intake and weight gain were collected at two nurseries (A, n=22; B, n=78) with the same feeding regimen protocols. Two strategies were used to explore similarity of regimen effect on caloric intake and weight gain. Repeated measures analysis of variance (ANOVA) was used to examine the effect on caloric intake and weight gain of time, feeding regimen, and time-by-regimen interaction for each nursery. RESULTS: In both nurseries, regimen effects were reasonably consistent for caloric intake and weight gain. Caloric intake was lower across nurseries for infants fed ad lib. After accounting for caloric intake, the ad lib regimen did not affect weight gain. The time-by-regimen interaction effect on caloric intake was significant in both nurseries. Caloric intake for infants fed ad lib increased significantly over 5 days. CONCLUSIONS: Despite differences between nurseries in infant characteristics and in protocol implementation, the feeding regimen effect was consistent for caloric intake and weight gain. Further support was found for the development of infant self-regulatory capacity.

Relationship between adolescent-parental communication and initiation of first intercourse by adolescents.

OBJECTIVE: To examine the level of communication between parents and adolescents and correlate the findings with onset of sexual intercourse. METHODS: This was a 10-year longitudinal study. Subjects filled out questionnaires at the time of each health supervision visit. The data in this article were harvested at enrollment and at the 5-year point. Comparisons of adolescent-parental communication were first compared between the virginal and nonvirginal groups at the initiation of the study. The results were then confirmed by adding data from patients in the virginal group who converted over the first 5 years of the study to the nonvirginal group. Two hundred and three patients, aged 12-21 years, were studied from a pediatrician's practice panel in Middleton, Wisconsin. The questionnaires inquired about grades and activities school, relationship with siblings and parents at home, and risk-taking behaviors involving alcohol, drugs, and sex with peers. Statistical analysis included the Student's t-test to determine the mean difference between groups. Fisher's exact test was used to evaluate the association of variables to the status of patients' sexual activity at enrollment. Multiple logistic regression was conducted on the initial enrollment data to examine the association between the initial covariates and patients' sexual status. RESULTS: During the initial survey, 172 enrolled patients were in the virginal group and 31 in the nonvirginal group. The virginal group had a higher rating of communication with their parents after adjusting for age (p <.001). To verify these findings, we examined an additional 29 patients in the virginal group who converted to nonvirginal status during the first 5 years of the study. We then compared the level of adolescent-parental communication between the subgroup who converted to nonvirginal status with the level of communication of the subgroup who remained virgins. Even after correcting for age, communication with the mother was significantly better in patients who maintained their virginal status (p <.01). CONCLUSIONS: Teenagers who perceive that they have a better level of communication with their parents are less likely to engage in sexual intercourse.

Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group.

OBJECTIVE: Despite its relative frequency among autosomal recessive diseases and the availability of the sweat test, cystic fibrosis (CF) has been difficult to diagnose in early childhood, and delays can lead to severe malnutrition, lung disease, or even death. The Wisconsin CF Neonatal Screening Project was designed as a randomized clinical trial to assess the benefits and risks of early diagnosis through screening. In addition, the incidence of CF was determined, and the validity of our randomization method assessed by comparing 16 demographic variables. METHODOLOGY: Immunoreactive trypsinogen analysis was applied to dried newborn blood specimens for recognition of CF risk from 1985 to 1991 and was coupled to DNA-based detection of the DeltaF508 mutation from 1991 to 1994. Randomization of 650 341 newborns occurred when their blood specimens reached the Wisconsin screening laboratory. This created 2 groups-an early diagnosis, screened cohort and a standard diagnosis or control group. To avoid selection bias, we devised a unique unblinding method with a surveillance program to completely identify the control subjects. Because sequential analysis of nutritional outcome measures revealed significantly better growth in screened patients during 1996, we accelerated the unblinding and completely identified the control group by April 1998. Having each member of this cohort enrolled and evaluated for at least 1 year and having completed a comprehensive surveillance program, we performed another statistical analysis of anthropometric evaluated indices that includes all CF patients without meconium ileus. RESULTS: The incidence of classical CF, ie, patients diagnosed in this trial with a sweat chloride of 60 mEq/L greater, was 1:4189. By incorporating other CF patients born during the randomization period, including 2 autopsy diagnosed patients and 8 probable patients, we calculate a maximum incidence of 1:3938 (95% confidence interval: 3402-4611). Although there were group differences in the proportion of patients with DeltaF508 genotypes and with pancreatic insufficiency, validity of the randomization plan was demonstrated by analyzing 16 demographic variables and finding no significant difference after adjustment for multiple comparisons. Focusing on patients without meconium ileus, we found a marked difference in the mean +/- standard deviation age of diagnosis for screened patients (13 +/- 37 weeks), compared with the standard diagnosis group (100 +/- 117). Anthropometric indices of nutritional status were significantly higher at diagnosis in the screened group, including length/height, weight, and head circumference. During 13 years of study, despite similar nutritional therapy and the inherently better pancreatic status of the control group, analysis of nutritional outcomes revealed significantly greater growth associated with early diagnosis. Most impressively, the screened group had a much lower proportion of patients with weight and height data below the 10th percentile throughout childhood. CONCLUSIONS: Although the screened group had a higher proportion of patients with pancreatic insufficiency, their growth indices were significantly better than those of the control group during the 13-year follow-up evaluation and, therefore, this randomized clinical trial of early CF diagnosis must be interpreted as unequivocally positive. Our conclusions did not change when the height and weight data before 4 years of age for the controls detected by unblinding were included in the analysis. Also, comparison of growth outcomes after 4 years of age in all subjects showed persistence of the significant differences. Therefore, selection bias has been eliminated as a potential explanation. In addition, the results show that severe malnutrition persists after delayed diagnosis of CF and that catch-up may not be possible. We conclude that early diagnosis of CF through neonatal screening combined with aggressive nutritional therapy can result

Masking unmasked in the proportional hazards model.

Influence measures based on the pairwise deletion approach and the differentiation approach are developed for unmasking observations masked by other observations in the proportional hazards model. These influential observations might have substantial impact on statistical inference and might provide important information for model adequacy. One numerical example based on real data is presented and discussed.

Wisconsin cystic fibrosis chest radiograph scoring system: validation and standardization for application to longitudinal studies.

This study was designed to achieve a final modeling, validation, and standardization plan for the Wisconsin cystic fibrosis (CF) chest radiographic scoring system. Sixty chest radiographs were selected to reflect a range of severity of lung pathology in children with CF. Seven experienced volunteer raters (three radiologists and four pediatric pulmonologists) from five institutions were recruited to evaluate and score the films. Analysis of scores revealed that the subcomponents of the Wisconsin system showed considerable variation from rater to rater, but reliability assessment indicated satisfactory Cronbach's alpha coefficients (0.83-0.90) among the seven raters. It was found that an additive method of total score computation is significantly more reliable (P < 0.05) than either the original multiplicative model or the traditional Brasfield scoring system. Comparison of radiologists and pulmonologists revealed a marked, systematic difference in scoring with the former group being more conservative in interpretation of abnormalities than the pulmonologists, and some of the raters showing very limited sensitivity. Quantitative chest radiology applied to children with cystic fibrosis studied long-term in longitudinal research projects requires the careful use of sensitive scoring methods and careful selection and training of multiple raters. This is particularly important since pulmonologists and radiologists can differ systematically in interpreting/scoring abnormalities.

Risk of persistent growth impairment after alternate-day prednisone treatment in children with cystic fibrosis.

BACKGROUND: It is uncertain whether the growth impairment that occurs in children during long-term treatment with glucocorticoids persists after the medication is discontinued and ultimately affects adult height. METHODS: We evaluated growth six to seven years after alternate-day treatment with prednisone had been discontinued in 224 children 6 to 14 years of age with cystic fibrosis who had participated in a multicenter trial of this therapy from 1986 through 1991. Of the children, 151 had been randomly assigned to receive prednisone (either 1 or 2 mg per kilogram of body weight) and 73 to receive placebo. We obtained data on growth up to 1997 from the Cystic Fibrosis Foundation Patient Registry and standardized the data to sex- and age-specific norms from the National Center for Health Statistics. We used z scores to compare growth patterns among treatment groups. RESULTS: In 1997, 68 percent of the patients were 18 years of age or older. The z scores for height declined during prednisone therapy; catch-up growth began two years after treatment with prednisone was discontinued. Among the boys, the z scores for height in those treated with prednisone remained lower than the scores for those who received placebo (P=0.02). The mean heights for boys 18 years of age or older were 4 cm less in the prednisone groups than in the placebo group, an equivalent of 13 percentile points (P=0.03). Among the girls, differences in height between those who were treated with prednisone and those who received placebo were no longer present two to three years after prednisone therapy was discontinued. CONCLUSIONS: Among children with cystic fibrosis who have received alternate-day treatment with prednisone, boys, but not girls, have persistent growth impairment after treatment is discontinued.

Nutritional status of patients with cystic fibrosis with meconium ileus: a comparison with patients without meconium ileus and diagnosed early through neonatal screening.

OBJECTIVE: This study was pursued as an extension of a randomized clinical investigation of neonatal screening for cystic fibrosis (CF). The objective was to determine if CF patients with meconium ileus (MI) were more likely to be malnourished compared with those without MI who were diagnosed during early infancy through neonatal screening. METHODOLOGY: Nutritional status was evaluated from early infancy to 13 years of age based on anthropometric, biochemical, and dietary assessments. RESULTS: MI patients (n = 32) were smaller at birth (3117 g compared with 3413 g) and were shorter (22nd percentile compared with 48th percentile) and thinner (24th percentile compared with 49th percentile) compared with non-MI early diagnosed patients (n = 50) up to 13 years of age. Poor growth was particularly evident in 26 MI patients who required surgery for MI (height and weight at the 20th percentile), whereas those treated without surgery (n = 6) showed better height (45th percentile) and weight (37th percentile). Abnormal essential fatty acid profiles were significantly more prevalent in MI compared with non-MI early-diagnosed patients before 3 years of age. Daily intakes of calorie (130% compared with 111% recommended dietary allowances) and protein (339% compared with 279% recommended dietary allowances) were higher but the percentage of fat (37% compared with 38%) and linoleic acid (4.5% compared with 4.7%) in the diet were similar between the two groups. CONCLUSIONS: These results demonstrated a clear association of MI with malnutrition in CF. The observed poor growth among our MI patients was not because of poor dietary intakes, but was related to surgical treatment for MI and poor essential fatty acid status. These findings present new challenges regarding the optimal medical treatment and nutritional intervention for CF patients with MI.

Pharmacokinetics of intraperitoneal epoetin alfa in patients on peritoneal dialysis using an 8-hour "dry dwell" dosing technique.

Pharmacokinetic studies of intraperitoneal (IP) epoetin alfa administered to continuous ambulatory peritoneal dialysis (CAPD) patients have shown low bioavailability, primarily attributable to the dilutional effect of coadministered dialysate. However, bioavailability is improved by instilling the dose into a dry peritoneum. The current study was designed to determine whether absorption after administration into a dry peritoneum is improved by extending the dry dosing period from 4 to 8 hours. The pharmacokinetics of a single 100-unit/kg IP epoetin alfa dose were studied in 8 noninfected CAPD patients. The dose was instilled into a dry peritoneum via the peritoneal catheter and allowed to dwell for 8 hours. CAPD was then resumed. Blood samples were collected for 96 hours after the dose. A 14-hour effluent dialysate sample was collected to determine epoetin alfa recovery. Enzyme immunoassay was used for epoetin alfa analysis of serum and effluent. Standard pharmacokinetic methods were employed for analysis of the serum concentration time data. The extent of epoetin alfa absorption was significantly greater than previously reported for a 4-hour dry dwell. The mean (+/-SD) dose-normalized area-under-the-curve (nlAUC(0-infinity)) using the 8-hour dry dwell dosing technique was 6,331 +/- 2,536 mIU. h/mL. This is significantly greater than the value of 2,589 +/- 1,450 mIU. h/mL (two-sided P value = 0.002) from a previous study in which patients received the same 100-unit/kg dose using a 4-hour dry dwell. The absorption of epoetin alfa administered by the intraperitoneal route is improved by extending the time the dose resides in a dry peritoneum.

Exact simultaneous confidence bands for a collection of univariate polynomials in regression analysis.

We discuss a simple simulation method for construction of exact confidence bands, having a pre-assigned confidence level simultaneously for several regression functions which are univariate polynomials in the predictors. This is accomplished by combining and extending existing results in a manner that permits both finite and infinite ranges for individual predictor variables. We illustrate the method for a logistic regression model with both dichotomous and continuous predictors.

The effects of prescribed versus ad libitum feedings and formula caloric density on premature infant dietary intake and weight gain.

BACKGROUND: Although feedings that are organized on an ad lib basis (i.e., in response to infant cues of hunger and of satiation) could enhance an infant's self-regulatory capacities for feeding, ad lib feeding of fully nipple-fed premature infants in a special care nursery has not been examined. OBJECTIVE: To study whether the caloric and protein intake and weight change of fully nipple-fed preterm infants differed by the feeding regimen (prescribed or ad lib) and by the caloric density of the formula (20- or 24-kcalories per ounce). METHOD: The 78 infants who participated in the study were randomized to prescribed or ad lib feeding regimens and, within each regimen, were further randomized to receive either 20-calorie or 24-kcalorie per ounce formula. Dietary intake (volume/kg, caloric intake/kg) and weight change (grams/kg gained or lost) were assessed for each of the 5 study days. Multivariate data analysis was used to examine the effects of feeding regimen and caloric density on dietary intake and weight change, controlling biologic variables (infant gender, race, lung disease diagnosis, treatment with supplemental oxygen, gestational age and weight at birth, and weight on the day prior to full nipple-feeding). RESULTS: Overall, the ad lib feeding regimen had a negative effect on volume intake and caloric intake. Weight gain was influenced by caloric intake, but not by feeding regimen or the caloric density of the diet. With increased full nipple-feeding experience, caloric intake of ad lib feeders approached that of the infants fed on the prescribed regimen. CONCLUSIONS: Development of self-regulatory capacities through ad lib feeding experience was indicated by infant regulation of the volume of intake by the caloric density of the formula, an unexpected finding. Furthermore, the approach of the caloric intake of infants on the ad lib regimen to that of infants on the prescribed regimen suggests they had gained skill in regulating intake with experience. Whether or not the trend for similar intakes would continue beyond 5 days is a question for further study.

Comparison of growth status of patients with cystic fibrosis between the United States and Canada.

BACKGROUND: Differences in growth status of patients with cystic fibrosis (CF) between the United States and Canada were reported in the 1980s based on analysis of data from 2 regional CF centers. OBJECTIVE: We evaluated the current growth status of the entire CF population in the United States and Canada in view of recent advances in the treatment of CF. DESIGN: Growth data from the 1992-1994 CF Patient Registries were analyzed. RESULTS: Mean height and weight were at approximately the 30th percentile for children with CF in the United States. Mean height and weight were 4-5 percentiles higher in children with CF in Canada than in those in the United States (P < 0.01), but percentages of ideal weight (104%) were similar in both populations. In adults with CF, mean height was similar at the 37th percentile; however, weight (26th compared with the 21st percentiles) and percentage of ideal weight (93% compared with 90%) were significantly higher in Canada than in the United States. Differences related to sex and age were similar in both countries for all indexes, which showed a high prevalence of underweight in infants and in older patients, but little sex discrepancy. CONCLUSION: We observed substantially smaller differences in the growth indexes of CF patients between the United States and Canada compared with results from the 1980s. These findings reflect significant improvements in the nutritional status of US patients in recent years. However, caution is required in the direct comparison of mean percentiles from reports using different growth standards because there are systematic differences in growth standards, which affect, in particular, the comparison of growth in males and females.

Neonatal jaundice and diet.

OBJECTIVE: To determine whether an earlier observation, that infants fed a casein-hydrolysate formula (Nutramigen) have lower neonatal jaundice levels than those fed standard formulas, would be repeated in a larger independent group of infants with more frequent measurements and more rigorous statistical analysis. DESIGN: Newborn infants were fed human milk, a standard whey-predominant formula (Enfamil), or Nutramigen (n = 20 for each group) during the first 3 weeks of life. Transcutaneous jaundice index was measured daily for the first week of life and every 2 to 3 days thereafter, using a noninvasive jaundice meter. Linear regression models of the data were constructed, validated, and compared statistically. SETTING: General community hospital with subsequent home visitation. PARTICIPANTS: Healthy, term newborn infants selected by convenience, based on time of birth. INTERVENTION: Infants were exclusively fed human milk, Enfamil, or Nutramigen. Formulas were randomly assigned. MAIN OUTCOME MEASURE: Jaundice index, a transcutaneous measurement of jaundice. RESULTS: The jaundice index differed significantly among the 3 groups. Paired comparisons showed that the jaundice index of the Nutramigen group was significantly lower than that of the Enfamil group (on days 6-16) and the human milk group (on days 3-20). The jaundice index of the Enfamil-fed group was significantly lower than that of the human milk group on days 13 to 19. CONCLUSIONS: Jaundice levels are lower in neonates fed Nutramigen rather than Enfamil and both these groups have lower jaundice levels than breast-fed infants.

Comparison of intraperitoneal and subcutaneous epoetin alfa in peritoneal dialysis patients.

OBJECTIVE: To compare the efficacy of intraperitoneal (i.p.) and subcutaneous (s.c.) administration of epoetin alfa in patients receiving peritoneal dialysis (PD). DESIGN: A 32-week prospective, randomized, cross-over experimental design. SETTING: Two university-based outpatient PD centers. PATIENTS: Twenty adult PD patients receiving stable doses of s.c. epoetin alfa enrolled in the study. Thirteen patients completed 32 weeks of follow-up. INTERVENTION: Patients were randomly assigned to receive either s.c. or i.p. epoetin alfa at the start of the study. Dose adjustments were made to maintain baseline hematocrit +/- 3 percentage points. Following 16 weeks of treatment, patients crossed over to the other route of administration for an additional 16 weeks. Intraperitoneal epoetin alfa was administered into an empty peritoneal cavity for approximately 8 hours before resuming dialysis. End-of-study i.p. epoetin alfa doses required to maintain target hematocrit were given twice weekly (n = 1), once weekly (n = 11), or once every other week (n = 1). All patients received iron supplements to maintain or exceed prestudy iron parameters. MAIN OUTCOME MEASURE: Prior to the study, the primary outcome measure was defined as the difference in epoetin alfa dose between i.p. and s.c. administration. RESULTS: Thirteen patients completed the study. The area under the dosing-requirement curve for i.p. epoetin alfa was larger than for s.c. administration (p = 0.0029), and the slope of the 16-week dose-requirement curve was greater for i.p. administration (p = 0.017), suggesting greater dose stability for s.c. administration. Paired analysis indicated greater i.p. intrapatient dose requirements (p < 0.0001). The mean difference in s.c. versus i.p. doses was 5000 +/- 1510 units per week. Some patients required escalating i.p. doses to maintain target hematocrit values. Iron administration and iron stores were similar in both groups. CONCLUSION: Intraperitoneal epoetin alfa may be a suitable alternative for some patients for whom s.c. dosing is undesirable. Large i.p. versus s.c. dosing differences noted in a few patients are unexplained, but may result from interpatient variability in i.p. epoetin alfa absorption. Intraperitoneal dosing into an empty peritoneum can be done safely and effectively.

Comprehensive analysis of risk factors for acquisition of Pseudomonas aeruginosa in young children with cystic fibrosis.

The objective of this study was to identify risk factors of significance for acquisition of Pseudomonas aeruginosa by children with cystic fibrosis (CF). Our working hypothesis is that exposure of infants and young children with CF to older, infected patients increases their risk for acquiring this organism. A special opportunity arose to study this question in detail, as we have been performing a randomized clinical trial of neonatal screening for CF throughout the state of Wisconsin during the period of 1985-1994. Patients were selected for this study based on either early identification through screening or diagnosis by standard methods. A longitudinal protocol employed at Wisconsin's two CF Centers includes routine cultures of respiratory secretions and collection of clinical, demographic, and activity information on patients and their families. Previous observations in our trial revealed that one center at an old hospital in an urban location showed a significantly shorter time to acquisition of P. aeruginosa for CF patients followed there. To study the center effect further, we performed statistical analyses using survival curves and stepwise regression analysis of all life history covariates available. The results of these analyses showed that the statistically significant correlations involve the following risk factors: 1) center and old hospital (r=0.42); 2) center and original physician (r=0.61); 3) center and exposure to pseudomonas-positive patients (r=0.29); and 4) population density and urban location (r=0.49). The final statistical model demonstrated that increased risk due to aerosol use (odds ratio=3.45, P=0.014) and a protective effect associated with education of the mother (odds ratio=0.81, P=0.024) were the most significant factors for acquisition of P. aeruginosa. The previously observed center effect was confined to the 1985-1990 interval at the old hospital (odds ratio=4.43, P < 0.001). We conclude that multiple factors are involved in increasing the risk of young children with CF to acquire P. aeruginosa, and that the observed center effect can best be explained by a combination of factors. These results suggest that facilities and methods used to care for young children with CF can significantly influence their likelihood of acquiring pseudomonas in the respiratory tract.