A183 VARIATION IN HEALTH SERVICES UTILIZATION AND RISK OF SURGERY ACROSS CHILDREN WITH INFLAMMATORY BOWEL DISEASE: A MULTIPROVINCE COHORT STUDY

Background

Marked variation in access to care and health services utilization is a marker of variation in quality of care. With the rising incidence of pediatric inflammatory bowel disease (IBD), we must understand variation in access to and outcomes of care to improve quality.

Purpose

Describe variation in care for pediatric IBD treated in 4 Canadian provinces.

Method

Incident cases of IBD diagnosed in children <16y were identified from health administrative data in Alberta (AB), Manitoba, Nova Scotia, and Ontario (ON) using validated algorithms. Children were assigned to one of 8 centres of care using a hierarchical assessment of health services use within 6 months of diagnosis. Children treated by adult gastroenterologists or community-based pediatric gastroenterologists were excluded due to small sample size. Outcomes included IBD-related hospitalizations, emergency department (ED) visits (AB/ON only), and IBD-related abdominal surgery. Hospitalizations and ED visits were counted cumulatively from 6-60 months after diagnosis. The risk of first surgery was defined during the same 6-60 month period. Mixed-effects meta-analysis was used to pool results across centres. Heterogeneity among centres was quantified using I² (variation in pooled event rates between centres) and τ (standard deviation of the true event rates). R² quantified the residual heterogeneity in outcomes not attributable to among-province variation.

Result(s)

We identified 3777 incident cases of pediatric IBD, 2936 (78%) of which were treated at 8 pediatric centres. The number of hospitalizations was 0.67 (95% CI 0.56-0.79) per person with high between-centre heterogeneity (I² 84%, τ 0.1556). Provincial differences accounted for 93% of heterogeneity across centres (residual heterogeneity: I² 29%, τ 0.0412). Hospitalizations were less frequent in AB than other provinces (0.43 vs. 0.72-0.78). Children averaged 1.94 IBD-related ED visits, with significant heterogeneity (I² 99%, τ 1.33) with 99.7% of heterogeneity attributable to among-province differences (residual heterogeneity: I² 32%; τ 0.074). Mean ED visits were 1.1 visits in ON (I² 39%) and 3.7 in AB (I² 0%). Intestinal resection was required by 12% (95% CI 0.08-0.15) of Crohn’s patients with high among-centre heterogeneity (I² 81%, τ 0.042), and low (19%) heterogeneity due to provincial differences (residual heterogeneity: I² 76%; τ 0.039). Colectomy was required by 12% (95% CI 10-14) of children with ulcerative colitis (UC) with no between-centre heterogeneity (I² 0%, τ 0).

Conclusion(s)

There is a high degree of between-province (but not between-centre, within province) variability in health services utilization among children with IBD. There was significant between-centre variability in surgery rates for Crohn’s, but not colectomy for UC. Differences in patient characteristics or provincial health systems may be more important predictors of variation in care. Surgery for Crohn’s disease may be a target for inter-centre quality improvement efforts.

Please acknowledge all funding agencies by checking the applicable boxes below

CCC

Disclosure of Interest

None Declared

A189 EMERGENCY DEPARTMENT UTILIZATION AND RISK OF INTESTINAL RESECTION IS LOWER AMONG CHILDREN DIAGNOSED WITH INFLAMMATORY BOWEL DISEASE BEFORE 10 YEARS OF AGE: A MULTIPROVINCE POPULATION-BASED COHORT STUDY

Background

In Canada, the incidence of inflammatory bowel disease (IBD) is increasing faster among those <10 years (y) of age than in any other age group. Understanding the health services burden of IBD in this population is important for health system planning.

Purpose

To compare healthcare utilization and risk of surgery among children diagnosed with IBD across age groups defined by the Paris Classification (A1a: <10y; A1b: 10 to <16y) across 5 Canadian provinces.

Method

Children diagnosed with IBD <16 years of age were identified from health administrative data using validated algorithms in Alberta, Manitoba, Nova Scotia, Ontario, and Québec. Negative binomial regression models were used to compare (1) the pre-diagnosis frequency of health services utilization (outpatient, emergency department (ED), and hospitalization) using diagnostic codes suggestive of future IBD and (2) the annual post-diagnosis frequency of IBD-specific and IBD-related (signs, symptoms, and extra-intestinal manifestations of IBD) visits among children diagnosed <10y (A1a) and 10 to <16y (A1b). Cox proportional hazard models compared the risk of surgery (identified with validated procedure codes) across age groups. All regression models were adjusted for sex, rural/urban residence, and mean neighbourhood income quintile. Province-specific event counts (all ages combined) and models (comparing age groups; reference: A1b [10 to <16y]) were pooled using random-effects meta-analysis.

Result(s)

Among 5124 children with IBD (1165 [23%] were <10y at diagnosis), the mean number of pre-diagnosis healthcare encounters was 1.0 (95% CI 0.38 to 1.68, I²=99.6%). The mean annual post-diagnosis number of IBD-specific outpatient visits was 3.2 (95% CI 1.9-4.4, I²=99.6%); hospitalizations, 0.19 (95% CI 0.17-0.21, I²=74%); ED visits, 0.17 (95% CI 0.19-0.39, I²=99%). The mean annual post-diagnosis number of IBD-related outpatient visits was 3.9 (95% CI 2.3-5.5, I²=99.7%); hospitalizations, 0.21 (95% CI 0.19-0.23, I²=79%); ED visits, 0.29 (95% CI 0.19-0.39, I²=97%). Intestinal resection or colectomy within 5y of diagnosis occurred in 13% (95%CI 8-22, I²=93%) with Crohn’s disease (CD) and 16% (95% CI 14-18, I²=40%) with ulcerative colitis. IBD-specific ED visits (RR 0.70, 95% CI 0.50-0.97, I²=80) and the risk of intestinal resection in CD (HR 0.49, 95% CI 0.26-0.92, I²=40%) were significantly lower among children diagnosed <10y. There were no age-related differences in pre-diagnosis health services utilization or other post-diagnosis outcomes, including frequency of outpatient visits to a gastroenterologist.

Conclusion(s)

Health services utilization was generally similar for children diagnosed with IBD at <10y and between 10 and <16y, except for lower rates of IBD-specific ED visits and intestinal resection in children with CD. Further exploration of between-province differences, represented by the high statistical heterogeneity (I²) in the meta-analyses, is needed to understand sources of variation in care.

Please acknowledge all funding agencies by checking the applicable boxes below

CCC

Disclosure of Interest

E. Kuenzig: None Declared, H. Singh Consultant of: Amgen Canada, Bristol-Myers Squibb Canada, Sandoz Canada, Roche Canada, Takeda Canada and Guardant Health, A. Bitton: None Declared, G. Kaplan Grant / Research support from: Ferring, Consultant of: AbbVie, Janssen, Pfizer, Amgen, Sandoz, Pendophram, and Takeda, Speakers bureau of: AbbVie, Janssen, Pfizer, Amgen, Sandoz, Pendophram, and Takeda, M. Carroll: None Declared, A. Otley Grant / Research support from: Research support: AbbVie Global. Research site: AbbVie, Pfizer, Eli-Lily, Janssen, Consultant of: AbbVie Canada, T. Stukel: None Declared, S. Spruin: None Declared, A. Griffiths Grant / Research support from: Abbvie, Consultant of: Abbvie, Amgen, BristolMyersSquibb, Janssen, Lilly, Takeda, Speakers bureau of: Abbvie, Janssen, Takeda, D. Mack: None Declared, K. Jacobson Grant / Research support from: Abbvie Canada and Janssen Canada, Consultant of: Abbvie Canada, Janssen Canada, Merck Canada and Mylan Pharmaceuticals, Speakers bureau of: Abbvie Canada and Janssen Canada, G. Nguyen: None Declared, L. Targownik Grant / Research support from: Janssen Canada, Consultant of: AbbVie Canada, Sandoz Canada, Takeda Canada, Merck Canada, Pfizer Canada, Janssen Canada, and Roche Canada, W. El-Matary Consultant of: Abbvie and MERCK, Speakers bureau of: Abbvie and MERCK, E. Benchimol Consultant of: McKesson Canada, Dairy Farmers of Ontario (unrelated to medications used to treat inflammatory bowel disease)

A26 PEDIATRIC-ONSET INFLAMMATORY BOWEL DISEASE INCREASES THE RISK OF VENOUS THROMBOEMBOLISM: A CANGIEC POPULATION-BASED STUDY

Background

Inflammatory bowel disease (IBD) increases the risk of venous thromboembolism (VTE) in patients of all ages but the risk of VTE among Canadian children with IBD has not previously been investigated.

Aims

Report the incidence of VTE and subtypes pulmonary embolism (PE) and deep vein thrombosis (DVT) in children with and without IBD.

Methods

Children diagnosed with IBD &lt;16y were identified from health administrative data in Ontario (2002–2014), Alberta (2007–2015), and Nova Scotia (2002–2012) using validated algorithms and matched by age and sex to children without IBD (1:5 ratio). Validated ICD-10 codes identified hospitalizations for incident VTE (DVT, PE, and sinovenous thrombosis). Province-specific 5-year cumulative incidence per 1000 person-years (PY) of VTEs were pooled using fixed-effects generalized linear mixed models with a Freeman-Tukey double arcsine transformation. Incidence rate ratios (IRR) within 5 years of diagnosis were pooled using fixed-effects generalized linear mixed models to compare children with and without IBD, and children with Crohn’s disease (CD) and ulcerative colitis (UC).

Results

3127 children with IBD (1826 CD; 1045 UC) were matched to 15,635 children without IBD. The cumulative incidence of VTE within 5 years of IBD diagnosis was 2.8 (95% CI 2.1–3.8) per 1000 PYs compared to 0.13 (95% CI 0.07–0.24) per 1000 PYs in children without IBD (Table). The 5-year cumulative incidences of VTE, DVT, and PE were significantly higher in children with IBD than in children without IBD (VTE: IRR 21.44, 95% CI 10.73–42.82; DVT: IRR 25.15, 95% CI 11.12–56.89; PE: IRR 4.01, 95% CI 1.22–13.18). Compared to UC patients, children with CD were at lower risk of VTE (IRR 0.53, 95% CI 0.29–0.96) and numerically, but not statistically, lower risk of DVT (IRR 0.59, 95% CI 0.30–1.14).

Conclusions

Although VTEs are relatively rare among children with IBD, these children are at much greater risk than children without IBD. Gastroenterologists caring for these patients should be cognizant of VTE risk and provide appropriate prophylaxis to those at high risk of VTE.

Funding Agencies

CCC

Impact of diabetes on healthcare costs in a population-based cohort: a cost analysis

AIMS

To estimate the healthcare costs attributable to diabetes in Ontario, Canada using a propensity-matched control design and health administrative data from the perspective of a single-payer healthcare system.

METHODS

Incident diabetes cases among adults in Ontario were identified from the Ontario Diabetes Database between 2004 and 2012 and matched 1:3 to control subjects without diabetes identified in health administrative databases on the basis of sociodemographics and propensity score. Using a comprehensive source of administrative databases, direct per-person costs (Canadian dollars 2012) were calculated. A cost analysis was performed to calculate the attributable costs of diabetes; i.e. the difference of costs between patients with diabetes and control subjects without diabetes.

RESULTS

The study sample included 699 042 incident diabetes cases. The costs attributable to diabetes were greatest in the year after diagnosis [C$3,785 (95% CI 3708, 3862) per person for women and C$3,826 (95% CI 3751, 3901) for men], increasing substantially for older age groups and patients who died during follow-up. After accounting for baseline comorbidities, attributable costs were primarily incurred through inpatient acute hospitalizations, physician visits and prescription medications and assistive devices.

CONCLUSIONS

The excess healthcare costs attributable to diabetes are substantial and pose a significant clinical and public health challenge. This burden is an important consideration for decision-makers, particularly given increasing concern over the sustainability of the healthcare system, aging population structure and increasing prevalence of diabetic risk factors, such as obesity.

The association of hospital prevention processes and patient risk factors with the risk of Clostridium difficile infection: a population-based cohort study

BACKGROUND

Clostridium difficile is the most common cause of healthcare-acquired infection; the real-world impacts of some proposed C. difficile prevention processes are unknown.

METHODS

We conducted a population-based retrospective cohort study of all patients admitted to acute care hospitals between April 2011 and March 2012 in Ontario, Canada. Hospital prevention practices were determined by survey of infection control programmes; responses were linked to patient-level risk factors and C. difficile outcomes in Ontario administrative databases. Multivariable generalised estimating equation (GEE) regression models were used to assess the impact of selected understudied hospital prevention processes on the patient-level risk of C. difficile infection, accounting for patient risk factors, baseline C. difficile rates and structural hospital characteristics.

RESULTS

C. difficile infections complicated 2341 of 653 896 admissions (3.6 per 1000 admissions). Implementation of the selected C. difficile prevention practices was variable across the 159 hospitals with isolation of all patients at onset of diarrhoea reported by 43 (27%), auditing of antibiotic stewardship compliance by 26 (16%), auditing of cleaning practices by 115 (72%), on-site diagnostic testing by 74 (47%), vancomycin as first-line treatment by 24 (15%) and reporting rates to senior leadership by 52 (33%). None of these processes were associated with a significantly reduced risk of C. difficile after adjustment for baseline C. difficile rates, structural hospital characteristics and patient-level factors. Patient-level factors were strongly associated with C. difficile risk, including age, comorbidities, non-elective and medical admissions.

CONCLUSIONS

In the largest study to date, selected hospital prevention strategies were not associated with a statistically significant reduction in patients' risk of C. difficile infection. These prevention strategies have either limited effectiveness or were ineffectively implemented during the study period.

Estimating gestational age at birth: a population-based derivation-validation study

Introduction

Information on newborn gestational age (GA) is essential in research on perinatal and infant health, but it is not always available from administrative databases. We developed and validated a GA prediction model for singleton births for use in epidemiological studies.

Methods

Derivation of estimated GA was calculated based on 130 328 newborn infants born in Ontario hospitals between 2007 and 2009, using linear regression analysis, with several infant and maternal characteristics as the predictor (independent) variables. The model was validated in a separate sample of 130 329 newborns.

Results

The discriminative ability of the linear model based on infant birth weight and sex was reasonably approximate for infants born before the 37th week of gestation (r2 = 0.67; 95% CI: 0.65–0.68), but not for term births (37–42 weeks; r2 = 0.12; 95% CI: 0.12–0.13). Adding other infant and maternal characteristics did not improve the model discrimination.

Conclusion

Newborn gestational age before 37 weeks can be reasonably approximated using locally available data on birth weight and sex.

Estimating gestational age at birth: a population-based derivation-validation study

INTRODUCTION

Information on newborn gestational age (GA) is essential in research on perinatal and infant health, but it is not always available from administrative databases. We developed and validated a GA prediction model for singleton births for use in epidemiological studies.

METHODS

Derivation of estimated GA was calculated based on 130 328 newborn infants born in Ontario hospitals between 2007 and 2009, using linear regression analysis, with several infant and maternal characteristics as the predictor (independent) variables. The model was validated in a separate sample of 130 329 newborns.

RESULTS

The discriminative ability of the linear model based on infant birth weight and sex was reasonably approximate for infants born before the 37th week of gestation (r2 = 0.67; 95% CI: 0.65-0.68), but not for term births (37-42 weeks; r2 = 0.12; 95% CI: 0.12-0.13). Adding other infant and maternal characteristics did not improve the model discrimination.

CONCLUSION

Newborn gestational age before 37 weeks can be reasonably approximated using locally available data on birth weight and sex.

Prostate cancer screening, treatment, and outcomes— A population-based study

10017

Background: It remains unknown whether more intense prostate cancer screening and treatment reduces prostate cancer mortality. We describe the experiences of two population-based cohorts with 15-years follow-up (1987–2001) to address the following questions: 1) does more intense screening and treatment for prostate cancer lead to lower mortality in community settings, and 2) do older men (age ≥75–79) benefit from more intense prostate cancer screening and treatment. Methods: Population-based cohort study consisting of white male Medicare beneficiaries who resided in the regions covered by the Seattle (N=88,863) and Connecticut (N=114,785) cancer registries. Inclusion criteria are age 65–79 and free of prostate cancer on January 1, 1987. All study subjects were followed through death or the end of 2001. The main outcomes are rates of screening for prostate cancer, treatment with radical prostatectomy, external beam radiotherapy, and prostate cancer specific mortality. Results: Between 1987 and 1990, compared to men in Connecticut, men in the Seattle region were 5.4 times (95% C.I. 4.8 - 6.1) more likely to undergo PSA testing, 2.2 times (95% C.I. 1.8 - 2.7) more likely to under go prostate biopsy, 5.9 times (95% C.I. 5.5 - 6.9) more likely to have radial prostatectomy, and 2.3 times (95% C.I. 2.2 - 2.5) more likely to have external beam radiation. The cumulative risk of radical prostatectomy or external beam radiation reached 9.1% in the Seattle cohort and 5.0% in the Connecticut cohort in 2001. After 15 years of follow-up, prostate cancer mortality rates were similar for subjects in the two study regions (hazard ratio of Seattle to CT: 1.01, 95% C.I. 0.93 - 1.09). For older men (aged 75–79 in 1987), however, the prostate cancer mortality rate was slightly higher in the Seattle than the Connecticut cohort (hazard ratio: 1.16, 95% C.I. 1.02 - 1.32). Conclusion: More intense screening for prostate cancer, surgery or radiation among a cohort of Medicare beneficiaries in the Seattle area compared with their counterparts in Connecticut has not lead to significantly lower mortality from prostate cancer over 15 years of follow-up.

No significant financial relationships to disclose.

A pooled analysis of 10 case-control studies of melanoma and oral contraceptive use

Data regarding the effects of oral contraceptive use on women's risk of melanoma have been difficult to resolve. We undertook a pooled analysis of all case-control studies of melanoma in women completed as of July 1994 for which electronic data were available on oral contraceptive use along with other melanoma risk factors such as hair colour, sun sensitivity, family history of melanoma and sun exposure. Using the original data from each investigation (a total of 2391 cases and 3199 controls), we combined the study-specific odds ratios and standard errors to obtain a pooled estimate that incorporates inter-study heterogeneity. Overall, we observed no excess risk associated with oral contraceptive use for 1 year or longer compared to never use or use for less than 1 year (pooled odds ratio (pOR)=0.86; 95% CI=0.74-1.01), and there was no evidence of heterogeneity between studies. We found no relation between melanoma incidence and duration of oral contraceptive use, age began, year of use, years since first use or last use, or specifically current oral contraceptive use. In aggregate, our findings do not suggest a major role of oral contraceptive use on women's risk of melanoma.

Are neonatal intensive care resources located according to need? Regional variation in neonatologists, beds, and low birth weight newborns

OBJECTIVE

Despite marked growth in neonatal intensive care during the past 30 years, it is not known if neonatologists and beds are preferentially located in regions with greater newborn risk. This study reports the relationship between regional measures of intensive care capacity and low birth weight infants using newly developed market-based regions of neonatal intensive care.

DESIGN

Cross-sectional small-area analysis of 246 neonatal intensive care regions (NICRs).

DATA SOURCES

1996 American Medical Association and American Osteopathic Association masterfiles data of clinically active neonatologists; 1999 American Academy of Pediatrics Section on Perinatal Pediatrics survey of directors of neonatal intensive care units in the United States with 100% response rate; 1995 linked birth/death data.

RESULTS

The number of total births per neonatologist across NICRs ranged from 390 to 8197 (median: 1722) and the number of total births per intensive care bed ranged from 72 to 1319 (median: 317). The associations between capacity measures and low birth weight rates across NICRs were statistically significant but negligible (R(2): 0.04 for neonatologists; 0.05 for beds). NICRs in the quintile with the greatest neonatologist capacity (average of only 863 births per neonatologist) had very low birth weight (VLBW) rates of 1.5% while those in the quintile of lowest neonatologist capacity (average of 3718 births per neonatologist) had VLBW rates of 1.3%; a similar lack of meaningful difference in VLBW rates was noted across quintiles of intensive care bed capacity. Including midlevel providers and intermediate care beds to the analyses did not alter the findings.

CONCLUSIONS

Neonatal intensive care capacity is not preferentially located in regions with greater newborn need as measured by low birth weight rates. Whether greater capacity affords benefits to the newborns remains unknown.

Two-stage methods for the analysis of pooled data

Epidemiologic studies of disease often produce inconclusive or contradictory results due to small sample sizes or regional variations in the disease incidence or the exposures. To clarify these issues, researchers occasionally pool and reanalyse original data from several large studies. In this paper we explore the use of a two-stage random-effects model for analysing pooled case-control studies and undertake a thorough examination of bias in the pooled estimator under various conditions. The two-stage model analyses each study using the model appropriate to the design with study-specific confounders, and combines the individual study-specific adjusted log-odds ratios using a linear mixed-effects model; it is computationally simple and can incorporate study-level covariates and random effects. Simulations indicate that when the individual studies are large, two-stage methods produce nearly unbiased exposure estimates and standard errors of the exposure estimates from a generalized linear mixed model. By contrast, joint fixed-effects logistic regression produces attenuated exposure estimates and underestimates the standard error when heterogeneity is present. While bias in the pooled regression coefficient increases with interstudy heterogeneity for both models, it is much smaller using the two-stage model. In pooled analyses, where covariates may not be uniformly defined and coded across studies, and occasionally not measured in all studies, a joint model is often not feasible. The two-stage method is shown to be a simple, valid and practical method for the analysis of pooled binary data. The results are applied to a study of reproductive history and cutaneous melanoma risk in women using data from ten large case-control studies.

Panel-based pain management in primary care. a pilot study

Although pain is an extremely common symptom presenting to primary care physicians, it frequently is not optimally managed. The purpose of this feasibility study was to develop and pilot-test an efficient, rapid assessment and management approach for pain in busy community practices. The intervention utilized the Dartmouth COOP Clinical Improvement System (DCCIS) and a telephone-based, nurse-educator intervention. Patients from four primary care practices in rural New Hampshire and Vermont were screened by mail for the presence of persistent pain. Patients with mild to severe pain were randomized to either the usual care control group (n = 383) or the intervention group (n = 320). Patients who reported pain but no psychosocial problems received a summary of identified problems and targeted educational material via mail (DCCIS). Patients who reported pain and psychosocial problems received the DCCIS intervention and calls from a nurse-educator who provided pain self-management strategies and a problem-solving approach for psychosocial problems. Post-treatment evaluation revealed that patients in the intervention group scored significantly better on the Pain, Physical, Emotional, and Social subscales of the SF-36 and on the total score of the Functional Interference Scale, as compared to a usual care control group. Feasibility and acceptability of the approach were demonstrated; however, the conclusions based on analyses of the post-treatment outcomes were tempered by baseline imbalances across groups.

The uneven landscape of newborn intensive care services: variation in the neonatology workforce

CONTEXT

In the past 30 years, the number of neonatologists has increased while total births have remained nearly constant. It is not known how equitably this expanded workforce is distributed.

OBJECTIVE

To determine the geographic distribution of neonatologists in the United States.

DATA SOURCES

1996 American Medical Association physician masterfiles; 1999 survey of all U.S. neonatal intensive care units; 1995 American Hospital Association hospital survey; and 1995 U.S. vital records.

MEASURES

The number of neonatologists and neonatal mid-level providers per live birth within 246 market-based regions.

RESULTS

The neonatology workforce varied substantially across neonatal intensive care regions. The number of neonatologists per 10,000 live births ranged from 1.2 to 25.6 with an interquintile range of 3.5 to 8.5. The weakly positive correlation between neonatologists and neonatal mid-level providers per live birth is not consistent with substitution of neonatal mid-level providers for neonatologists (Spearman rank-correlation coefficient, 0.17; P < 0.01). There was no difference in the percentage of neonatal fellows in the lowest and highest workforce quintile (14% vs. 16%) or in the percentage of neonatologists engaged predominantly in research, teaching, or administration (14% in lowest and highest quintiles).

CONCLUSIONS

The regional supply of neonatologists varies dramatically and cannot be explained by the substitution of neonatal mid-level providers or by the presence of academic medical centers. Further research is warranted to understand whether neonatal intensive care resources are located in accordance with risk and whether more resources improve newborn outcomes.

Skin cancer risk in relation to toenail arsenic concentrations in a US population-based case-control study

Arsenic is a known carcinogen specifically linked to skin cancer occurrence in regions with highly contaminated drinking water or in individuals who took arsenic-containing medicines. Presently, it is unknown whether such effects occur at environmental levels found in the United States. To address this question, the authors used data collected on 587 basal cell and 284 squamous cell skin cancer cases and 524 controls interviewed as part of a case-control study conducted in New Hampshire between 1993 and 1996. Arsenic was determined in toenail clippings using instrumental neutron activation analysis. The odds ratios for squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) were close to unity in all but the highest category. Among individuals with toenail arsenic concentrations above the 97th percentile, the adjusted odds ratios were 2.07 (95% confidence interval (CI): 0.92, 4.66) for SCC and 1.44 (95% CI: 0.74, 2.81) for BCC, compared with those with concentrations at or below the median. While the risks of SCC and BCC did not appear elevated at the toenail arsenic concentrations detected in most study subjects, the authors cannot exclude the possibility of a dose-related increase at the highest levels of exposure experienced in the New Hampshire population.

Therapeutic ionizing radiation and the incidence of basal cell carcinoma and squamous cell carcinoma. The New Hampshire Skin Cancer Study Group

OBJECTIVE

To estimate the relative risk of developing basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) after receiving therapeutic ionizing radiation.

DESIGN

Population-based case-control study.

SETTING

New Hampshire.

PATIENTS

A total of 592 cases of BCC and 289 cases of SCC identified through a statewide surveillance system and 536 age- and sex-matched controls selected from population lists.

MAIN OUTCOME MEASURES

Histologically confirmed BCC and invasive SCC diagnosed between July 1, 1993, through June 30, 1995, among New Hampshire residents.

RESULTS

Information regarding radiotherapy and other factors was obtained through personal interviews. An attempt was made to review the radiation treatment records of subjects who reported a history of radiotherapy. Overall, an increased risk of both BCC and SCC was found in relation to therapeutic ionizing radiation. Elevated risks were confined to the site of radiation exposure (BCC odds ratio, 3. 30; 95% confidence interval, 1.60-6.81; SCC odds ratio, 2.94; 95% confidence interval, 1.30-6.67) and were most pronounced for those irradiated for acne exposure. For SCC, an association with radiotherapy was observed only among those whose skin was likely to sunburn with sun exposure.

CONCLUSIONS

These results largely agree with those of previous studies on the risk of BCC in relation to ionizing radiation exposure. In addition, they suggest that the risk of SCC may be increased by radiotherapy, especially in individuals prone to sunburn with sun exposure. Arch Dermatol. 2000;136:1007-1011

Associations among hospital capacity, utilization, and mortality of US Medicare beneficiaries, controlling for sociodemographic factors

OBJECTIVE

To explore whether geographic variations in Medicare hospital utilization rates are due to differences in local hospital capacity, after controlling for socioeconomic status and disease burden, and to determine whether greater hospital capacity is associated with lower Medicare mortality rates.

DATA SOURCES/STUDY SETTING

The study population: a 20 percent sample of 1989 Medicare enrollees. Measures of resources were based on a national small area analysis of 313 Hospital Referral Regions (HRR). Demographic and socioeconomic data were obtained from the 1990 U.S. Census. Measures of local disease burden were developed using Medicare claims files.

STUDY DESIGN

The study was a cross-sectional analysis of the relationship between per capita measures of hospital resources in each region and hospital utilization and mortality rates among Medicare enrollees. Regression techniques were used to control for differences in sociodemographic characteristics and disease burden across areas.

DATA COLLECTION/EXTRACTION METHODS

Data on the study population were obtained from Medicare enrollment (Denominator File) and hospital claims files (MedPAR) and U.S. Census files.

PRINCIPAL FINDINGS

The per capita supply of hospital beds varied by more than twofold across U.S. regions. Residents of areas with more beds were up to 30 percent more likely to be hospitalized, controlling for ecologic measures of socioeconomic characteristics and disease burden. A greater proportion of the population was hospitalized at least once during the year in areas with more beds; death was also more likely to take place in an inpatient setting. All effects were consistent across racial and income groups. Residence in areas with greater levels of hospital resources was not associated with a decreased risk of death.

CONCLUSIONS

Residence in areas of greater hospital capacity is associated with substantially increased use of the hospital, even after controlling for socioeconomic characteristics and illness burden. This increased use provides no detectable mortality benefit.

Has asthma medication use in children become more frequent, more appropriate, or both?

OBJECTIVE

Despite national initiatives to improve asthma medical treatment, the appropriateness of physician prescribing for children with asthma remains unknown. This study measures trends and recent patterns in the pediatric use of medications approved for reversible obstructive airway disease (asthma medications).

DESIGN

Population-based longitudinal and cross-sectional analyses. Setting. A nonprofit staff model health maintenance organization located in the Puget Sound area of Washington state.

PARTICIPANTS

Children 0 to 17 years of age enrolled continuously during any one of the years from 1984 to 1993 (N = 83 232 in 1993). PRIMARY OUTCOME MEASURES. Percent of enrollees filling prescriptions for asthma medications and fill rates by medication class and estimated duration of inhaled antiinflammatory medication use.

RESULTS

Between 1984 and 1993, the frequency of asthma medication use increased: the percent of children filling any asthma medication prescription increased from 4. 0% to 8.1%, whereas the percent filling an inhaled antiinflammatory inhaler rose from 0.4% to 2.4%. In contrast, the intensity of inhaled antiinflammatory use decreased among users; 37% of users filled more than two inhalers during the year in 1984, and 29% in 1993. In high beta-agonist users (filling more than two beta-agonist inhalers each quarter per year), the estimated duration of inhaled antiinflammatory use increased slightly from a mean of 4.1 months per year in 1984-1986 to 5.0 months in 1991-1993; estimated duration of use in adolescents 10 to 17 years of age was approximately half that of children 5 to 9 years of age.

CONCLUSIONS

The proportion of children using asthma medications increased substantially during the study period, but the use of inhaled antiinflammatory medication per patient remained low even for those using large amounts of inhaled beta-agonists. These findings suggest that most asthma medications were used by children with mild lower airway symptoms and that inhaled antiinflammatory medication use in children with more severe disease fell short of national guidelines.

Increase in incidence rates of basal cell and squamous cell skin cancer in New Hampshire, USA. New Hampshire Skin Cancer Study Group

We conducted a study to estimate the current incidence rates of basal-cell carcinoma (BCC) and squamous-cell carcinoma (SCC) of the skin in the population of New Hampshire (NH), USA, and to quantify recent changes in the incidence rates of these malignancies. BCCs and SCCs diagnosed among NH residents were identified through physician practices and central pathology laboratories in NH and bordering regions from June 1979 through May 1980 and from July 1993 through June 1994. For each diagnosis period, we estimated the age-adjusted incidence rates for both BCC and SCC among both men and women and for separate anatomic sites. Between 1979-1980 and 1993-1994, incidence rates of SCC increased by 235% in men and by 350% in women. Incidence rates of BCC increased by more than 80% in both men and women. While the absolute increase was greatest for tumors of the head and neck, the relative change was most pronounced for tumors on the trunk in men and on the lower limb in women. Thus, there has been a marked rise in the incidence rates of BCC and SCC skin cancers in NH in recent years. The anatomic pattern of increase in BCC and SCC incidence is consistent with an effect of greater sunlight exposure. Studies of BCC and SCC occurrence are needed to identify possible behavioral and environmental factors and to assess possible changes in diagnostic practices that might account for the rise in incidence of these common malignancies.

A randomized trial of the use of patient self-assessment data to improve community practices

OBJECTIVE

Because of time constraints in the office environment, problems of concern to elderly patients may not be raised during clinic visits. To facilitate communication about geriatric health problems, we examined the impact of a strategy that used patient self-assessment data to improve community practices.

DESIGN

Twenty-two primary care practices were randomized to participate in the intervention strategy (intervention practices) or to provide usual care (usual care practices).

SETTING

Primary care practices in 16 towns in New Hampshire (total, 45 physicians).

PATIENTS

1651 patients 70 years of age or older.

INTERVENTION

All patients received a mailed survey that asked about their health problems and about how well these problems were being addressed by their physicians. In the intervention practices, these data were used to generate a customized letter that directed the patient to specific sections in an 80-page modified version of the National Institute on Aging's Age Pages and were summarized and communicated to the patient's physician.

MAIN OUTCOME MEASURE

Change from baseline in patients' overall assessment of health care.

RESULTS

In 8 of 11 intervention practices, patients felt that their care had improved over the 2-year study period. This improvement occurred in only 1 of 11 usual care practices (P = 0.003). Patients in intervention practices reported receiving significantly more help with physical function, fall prevention, and assistance for memory problems. Self-assessed health status did not differ in the two groups.

CONCLUSION

A standard, easy-to-implement strategy to improve the quality of provider--patient interactions can improve the satisfaction of older patients cared for in community practices.

Physical symptom trajectories following trauma exposure: longitudinal findings from the normative aging study

This study modeled physical symptom trajectories from ages 30 to 75 in 1079 older male military veterans who were assessed every 3 to 5 years since the 1960s. Combat exposure and noncombat trauma were used to define four groups: no trauma (N = 249), noncombat trauma only (N = 333), combat only (N = 152), and both combat and noncombat trauma (N = 345). Number of symptoms on the Cornell Medical Index physical symptom scale increased 29% per decade. Men who had experienced either combat or noncombat trauma did not differ from nonexposed men, but those who had experienced both combat and noncombat trauma had 16% more symptoms across all ages. There were no differences in age-related trajectories as a function of trauma history. In cross-sectional analysis, men with combat and noncombat trauma had more posttraumatic stress disorder symptoms, but not more depression symptoms, than men with either no trauma or noncombat trauma only. Discussion focuses on the importance of considering physical as well as psychological outcomes of exposure to traumatic events.

Brain size, head size, and intelligence quotient in monozygotic twins

Many studies of monozygotic (MZ) twins have revealed evidence of genetic influences on intellectual functions and their derangement in certain neurologic and psychiatric diseases afflicting the forebrain. Relatively little is known about genetic influences on the size and shape of the human forebrain and its gross morphologic subdivisions. Using MRI and quantitative image analysis techniques, we examined neuroanatomic similarities in MZ twins and their relationship to head size and intelligence quotient (IQ). ANOVA were carried out using each measure as the dependent variable and genotype, birth order, and sex, separately, as between-subject factors. Pairwise correlations between measures were also computed. We found significant effects of genotype but not birth order for the following neuroanatomic measures: forebrain volume (raw, p < or = 0.0001; normalized by body weight, p = 0.0003); cortical surface area (raw, p = 0.002; normalized, p = 0.001); and callosal area (raw, p < or = 0.0001; normalized by forebrain volume, p = 0.02). We also found significant effects of genotype but not birth order for head circumference (raw, p = 0.0002; normalized, p < or = 0.0001) and full-scale IQ (p = 0.001). There were no significant sex effects except for raw head circumference (p = 0.03). Significant correlations were observed among forebrain volume, cortical surface area, and callosal area and between each brain measure and head circumference. There was no significant correlation between IQ and any brain measure or head circumference. These results indicate that: 1) forebrain volume, cortical surface area, and callosal area are similar in MZ twins; and 2) these brain measures are tightly correlated with one another and with head circumference but not with IQ in young, healthy adults.

Trends in pediatric asthma hospitalization rates: regional and socioeconomic differences

OBJECTIVE

Asthma hospitalization rates continue to increase nationally for children despite efforts by the National Institutes of Health and specialty organizations to improve outcomes through the dissemination of practice guidelines. To understand the generalizability of national trends to regional populations, we studied childhood hospitalizations over a 10-year period in four northeastern states.

DESIGN

Longitudinal analysis of hospitalization rates by patient residence and patient characteristics using state hospital discharge datasets.

POPULATION

Age < 18 years residing in Maine, New Hampshire, Vermont, or New York state during the period 1985 to 1994.

RESULTS

In multivariate analyses (controlling for age, sex, race/ethnicity, median household income, metropolitan status), we found that New York asthma hospitalization rates increased 3.8% per annum (95% confidence interval: 3.3, 4.2), whereas in New Hampshire, rates decreased 5.8% (95% confidence interval: 7.6, 4.1). Maine and Vermont rates did not change significantly during the study period. Increased asthma hospitalization rates were noted in black and Hispanic populations, in children residing in zip codes with lower median household incomes, and in those living in metropolitan areas. Hospitalization rates for nonasthma causes fell substantially. As a result, the proportion of hospital days attributed to childhood asthma increased in all population groups.

CONCLUSIONS

Asthma discharge rates measured by the state of residence or socioeconomic characteristic do not necessarily parallel national trends. None of the current hypotheses offered to explain national trends in asthma hospitalization rates (changes in disease severity, diagnostic substitution, or differences in the supply and character of medical care) can be the sole explanation of these regional trends. Efforts intended to improve asthma outcomes may benefit a greater number of children by redirecting resources toward specific populations identified through state hospital discharge datasets.

A comparison of clomiphene citrate and human menopausal gonadotropin for use in conjunction with intrauterine insemination

OBJECTIVE

To compare the outcome of superovulation using clomiphene citrate (CC) versus hMG in conjunction with IUI.

DESIGN

Sequentially assigned, observational study. Couples initially were assigned to receive either CC or hMG for three cycles.

SETTING

The Clinical Outpatient Department of the Dartmouth-Hitchcock Medical Center.

PATIENT(S)

Eighty-three infertile couples.

INTERVENTION(S)

IUI with hMG use.

MAIN OUTCOME MEASURE(S)

Conception rate, term pregnancy rate (PR), and pregnancy complications, such as spontaneous miscarriage and multiple gestation.

RESULT(S)

Of 83 couples who underwent at least one treatment cycle, 29 (35%) conceived during the study period. The relative rate of conception for hMG versus CC was 2.08 (95% confidence interval [CI], 0.93 to 4.68). The relative term PR was 2.10 (95% CI, 0.77 to 5.73) for hMG versus CC. There was no difference in the miscarriage rate for hMG versus CC.

CONCLUSION(S)

Both the conception rate and the term PR were higher using hMG, compared with CC, in combination with IUI, and showed a trend toward statistical significance.

The distance to community medical care and the likelihood of hospitalization: is closer always better?

OBJECTIVES

This study examined the influence that distance from residence to the nearest hospital had on the likelihood of hospitalization and mortality.

METHODS

Hospitalizations were studied for Maine. New Hampshire, and Vermont during 1989 (adults) and for 1985 through 1989 (children) and for mortality (1989) in Medicare enrollees.

RESULTS

After other known predictors of hospitalization (age, sex, bed supply, median household income, rural residence, academic medical center, and presence of nursing home patients) were controlled for, the adjusted rate ratio of medical hospitalization for residents living more than 30 minutes away was 0.85 (95% confidence interval [CI] = 0.82, 0.88) for adults and 0.78 (95% CI = 0.74, 0.81) for children, compared with those living in a zip code with a hospital. Similar effects were seen for the four most common diagnosis-related groups for both adults and children. The likelihood of hospitalization for conditions usually requiring hospitalization and for mortality in the elderly did not differ by distance.

CONCLUSIONS

Distance to the hospital exerts an important influence on hospitalization rates that is unlikely to be explained by illness rates.

Two-stage method of estimation for general linear growth curve models

We extend the linear random-effects growth curve model (REGCM) (Laird and Ware, 1982, Biometrics 38, 963-974) to study the effects of population covariates on one or more characteristics of the growth curve when the characteristics are expressed as linear combinations of the growth curve parameters. This definition includes the actual growth curve parameters (the usual model) or any subset of these parameters. Such an analysis would be cumbersome using standard growth curve methods because it would require reparameterization of the original growth curve. We implement a two-stage method of estimation based on the two-stage growth curve model used to describe the response. The resulting generalized least squares (GLS) estimator for the population parameters is consistent, asymptotically efficient, and multivariate normal when the number of individuals is large. It is also robust to model misspecification in terms of bias and efficiency of the parameter estimates compared to maximum likelihood with the usual REGCM. We apply the method to a study of factors affecting the growth rate of salmonellae in a cubic growth model, a characteristic that cannot be analyzed easily using standard techniques.

Effects of anesthetic technique on myocardial wall motion abnormalities during abdominal aortic surgery

OBJECTIVE

To assess the impact of regional supplemented general anesthesia (RSGEN) on regional myocardial function during abdominal aortic surgery (AAS).

DESIGN

Prospective randomized study.

SETTING

Single academic medical center.

PARTICIPANTS

Seventy-three patients scheduled for infrarenal aortic aneursymectomy.

INTERVENTIONS

Patients received standardized intraoperative anesthetic management consisting of either general anesthesia (GA; n = 37) or general anesthesia supplemented by epidural anesthesia (RSGEN; n = 36).

MEASUREMENTS AND MAIN RESULTS

Hemodynamic measurements and transesophageal echocardiograms (TEE) were obtained at eight intraoperative times. The electrocardiogram (ECG) was continuously recorded using Holter monitoring. Of the 56 patients with interpretable TEE recordings, 8 of 30 (27%) GA patients and 7 of 26 (27%) RSGEN patients developed new segmental wall motion abnormalities (SWMAs). There was no treatment effect on either the incidence (p = 0.23) or the intensity (p = 0.34) of SWMAs. Cross-clamping of the aorta was associated with the onset of new SWMAs (odds ratio, 8.2; 95% CI, 1.1 to 64; p = 0.04). Among the 63 patients with interpretable Holter recordings, 9 of 34 (26%) GA patients and 9 of 29 (31%) RSGEN patients exhibited intraoperative ischemia. There was no treatment effect on the incidence (p = 0.22) or intensity (p = 0.67) of ECG ischemia.

CONCLUSION

Despite providing modest hemodynamic depression, RSGEN did not reduce the incidence or intensity of either regional myocardial dysfunction or ECG ischemia. New SWMAs were temporally associated with cross-clamping of the aorta and tended to resolve with unclamping.

Superficial femoral artery stenoses: characteristics of progressing lesions

PURPOSE

The purpose of this study was to assess the effect of superficial femoral artery (SFA) stenosis morphologic characteristics and lesion location on the rate of atherosclerotic disease progression.

METHODS

We identified 19 patients who required arteriography for treatment of critical leg ischemia and who had previously undergone arteriography of that leg when minimal or no symptoms were present. These initial incidental arteriographic evaluations were performed during evaluation of arterial disease in another vascular bed from 4 to 81 months (mean, 32 months) previously. Distinct SFA stenoses or occlusion on the final arteriogram (n = 98) were characterized by their location, length, stenosis severity, and morphologic appearance on the initial arteriogram. The contribution of patient-specific risk factors to disease progression was also assessed.

RESULTS

Stenosis progression occurred independently among multiple lesions within the same patient (negligible intraclass correlation coefficient, r = 0.06). Lesions in the adductor canal region were more likely to occlude than lesions elsewhere in the SFA (adjusted odds ratio = 10.7; p = 0.03). Severity of initial lesion stenosis also was predictive of occlusion (adjusted odds ratio = 1.8; p = 0.04). However, most progressing lesions (93%) actually arose in areas of initially mild disease (stenoses < 50%) despite more severe initial lesions elsewhere. Increasing age (p = 0.023) and previous contralateral leg bypass (p = 0.036) were also associated with increasing rates of lesion progression. Smooth-asymmetric lesions progressed 11% more slowly than other lesion types (p = 0.003).

CONCLUSIONS

Our analysis of atherosclerotic SFA lesion progression in patients with critical ischemia shows that initial stenosis severity was associated with higher occlusion rates and that smooth-asymmetric lesions progressed more slowly than lesions with other morphologic characteristics. Severe stenoses usually arose from minimally diseased regions and progressed more rapidly than preexisting, more highly stenotic lesions. Most SFA occlusions resulted from disease progression in the adductor canal region whether or not antecedent lesions were seen on arteriography and whether or not more severe stenoses were initially present elsewhere. Increased age and history of previous contralateral bypass were patient-specific predictors of lesion progression.

Risk of squamous cell carcinoma of the skin in relation to plasma selenium, alpha-tocopherol, beta-carotene, and retinol: a nested case-control study

We conducted a nested case-control study of squamous cell skin cancer (SCC) to determine whether risk was related to plasma concentrations of selenium, alpha-tocopherol, beta-carotene, and retinol. We derived the study sample from participants in our Skin Cancer Prevention Study, all of whom had at least one basal cell or squamous cell skin cancer before study entry. Those who developed a new squamous cell skin cancer during the 3-5-year follow-up period were selected as cases (n = 132). Controls (n = 264) were chosen at random, with matching by age, sex, and study center, from among those who did not develop SCC but were being followed actively at the time the SCC case was diagnosed. Prediagnostic plasma samples were analyzed for alpha-tocopherol, beta-carotene, and retinol using high-performance liquid chromatography. Selenium determinations were made using instrumental neutron activation analysis. Odds ratios were computed using conditional logistic regression for matched samples. We found no consistent pattern of SCC risk associated with any of the nutrients examined. The odds ratios (95% confidence intervals) for the highest versus the lowest quartiles of beta-carotene, retinol, alpha-tocopherol, and selenium were 0.73 (0.38-1.41), 1.43 (0.77-2.64), 0.89 (0.43-1.85), and 0.86 (0.47-1.58), respectively. Thus, our data add to the growing body of evidence that these nutrients, at the concentrations we evaluated, are not related strongly to SCC risk.

Risk of basal cell and squamous cell skin cancers after ionizing radiation therapy. For The Skin Cancer Prevention Study Group

BACKGROUND

Human evidence that ionizing radiation is carcinogenic first came from reports of nonmelanoma skin cancers (NMSCs) on the hands of workers using early radiation devices. An increased risk of NMSC has been observed among uranium miners, radiologists, and individuals treated with x rays in childhood for tinea capitis (ringworm of the scalp) or for thymic enlargement; NMSC is one of the cancers most strongly associated with the atomic bombing of Hiroshima and Nagasaki. Although exposure to ionizing radiation is a known cause of NMSC, it is not yet clear whether therapeutic radiation causes both major histologic types of NMSC, basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). Additionally, the potentially modifying effects, such as latency, age when treated, and type of treatment, are not well understood.

PURPOSE

We investigated the relative risks of BCC and SCC associated with previous radiation therapy and evaluated these risks in relation to age and time since initial treatment and the medical condition for which radiation therapy was given.

METHODS

The study group comprised individual diagnosed with at least one BCC or SCC from January 1980 through February 1986, who were recruited to participate in a skin cancer prevention trial designed to test whether oral beta-carotene supplementation would reduce the risk of new NMSCs. Patients were identified through the dermatology and pathology records of academic medical centers in Hanover, NH; Los Angeles, CA; San Francisco, CA; and Minneapolis, MN. Each participant completed a questionnaire detailing lifetime residence, pigmentary characteristics, occupational and recreational sun exposure, and history of radiation therapy. At enrollment, a study dermatologist assessed skin type (tendency to burn or tan) and extent of actinic skin damage. Participants were followed with an annual dermatologic examination for an average of 4 years. Of the 5232 potentially eligible individuals, 1805 were enrolled in the trial. We excluded 112 patients who reported previous radiation therapy for skin cancer only and three with missing information on whether they were ever treated with radiation therapy, leaving 1690 patients for the analysis. Approximately 4% of the patients died or discontinued participation for other reasons during each study year. We examined time to occurrence of first new histopathologically confirmed BCC and SCC during the follow-up period in relation to history of radiation therapy (for reasons other than NMSC) using a proportional hazards model. A multiple end points survival model was used to compare the rate ratios (RRs) for BCC and SCC. We also used a longitudinal method of analysis to compute the RR of total new BCC and SCC tumors per person per study year associated with radiation therapy. Using this method, we additionally assessed the potential modifying effects of age at treatment, latency, and type of therapy. All P values were derived from two-sided statistical tests of significance.

RESULTS

Among the participants we studied, 597 developed a new BCC (n = 1553 tumors) and 118 developed a new SCC (n = 179 tumors). The time to first new BCC, but not SCC, was associated with prior radiation therapy (RR = 1.7; 95% confidence interval [CI] = 1.4-2.0 and RR = 1.0; 95% CI = 0.6-1.7, respectively; P = .03 for the difference between the RRs). The RR of total BCC tumors was slightly higher (RR = 2.3; 95% CI = 1.7-3.1), but it was still unity for SCC (RR = 1.0; 95% CI = 0.5-1.9). BCC risk appeared to increase with younger age at exposure and time since initially treated, although these effects were only marginally statistically significant (P for trend = .06 and .07, respectively). Also, risk of BCC was more strongly related to treatment for acne (RR = 3.3; 95% CI = 2.1-5.2) than other conditions.

CONCLUSIONS AND IMPLICATIONS

Our data suggest that exposure to therapeutic radiation is associated with BCC but not with SCC.

Iron deficiency in Massachusetts communities: Socioeconomic and demographic risk factors among children

OBJECTIVES

This study examined the association between community rates of iron deficiency in children and sociodemographic characteristics of Massachusetts communities.

METHODS

Between April 1990 and March 1991, 238 273 Mssachusetts children 6 through 59 months of age were screened; iron deficiency was defined as an erythrocyte protopophyrin concentration of 0.62 micromol/L or higher and a blood lead level of less than 1.2 micromol/L. Sociodemographic data were obtained from the 1990 US Census.

RESULTS

Five percent of communities had iron deficiency rates greater than 13.9 per 100 children screened. Iron deficiency rate was positively associated with proportion of Southeast Asians (odds ratio [OR] = 1.10, 95% confidence interval [CI] = 1.08, 1.12), proportion of Hispanics (OR = 1.008, 95% CI = 1.002, 1.013), and high school incompletion (OR = 1.028, 95% CI = 1.020, 1.035). Similarly, an examination of three Massachusetts cities indicated that the iron deficiency rate was higher for children with Southeast Asian (relative risk [RR] = 3.6, 95% CI = 3.3, 3.8) and Hispanic (RR = 1.6, 95% CI = 1.5, 1.8) surnames than for all other children.

CONCLUSIONS

Wide variation exists in iron deficiency rates for children in Massachusetts communities. Community iron deficiency was associated with low socioeconomic status and high proportions of Southeast Asians and Hispanics.

Mortality associated with low plasma concentration of beta carotene and the effect of oral supplementation

OBJECTIVE

To examine the relationship between beta carotene plasma concentration and beta carotene supplementation and risk of death from major disease causes.

DESIGN

Cohort study of plasma concentrations; randomized, controlled clinical trial of supplementation.

SETTING

Medical school-affiliated dermatology practices.

PATIENTS

A total of 1188 men and 532 women with mean age of 63.2 years, who had enrolled in a randomized clinical trial of beta carotene supplementation to prevent nonmelanoma skin cancer.

INTERVENTION

Oral beta carotene, 50 mg per day for a median of 4.3 years.

MAIN OUTCOME MEASURES

All-cause mortality and mortality from cardiovascular disease and cancer.

RESULTS

During a median follow-up period of 8.2 years, there were 285 deaths. Persons whose initial plasma beta carotene concentrations were in the highest quartile (>0.52 micromol/L [27.7 microg/dL]) had a lower risk of death from all causes (adjusted relative rate [RR], 0.52; 95% confidence interval [CI] 0.44 to 0.87) and from cardiovascular diseases (adjusted RR, 0.57; 95% CI, 0.34 to 0.95) compared with persons with initial concentrations in the lowest quartile (<0.21 micromol/L [11.2 microg/dL]). Patients randomly assigned to beta carotene supplementation showed no reduction in relative mortality rates from all causes (adjusted RR, 1.03; 95% CI, 0.82 to 1.30) or from cardiovascular disease (adjusted RR, 1.16; 95% CI, 0.82 to 1.64). There was no evidence of lower mortality following supplementation among patients with initial beta carotene concentrations below the median for the study group.

CONCLUSIONS

These analyses provide no support for a strong effect of supplemental beta carotene in reducing mortality from cardiovascular disease or other causes. Although the possibility exists that beta carotene supplementation produces benefits that are too small or too delayed to have been detected in this study, noncausal explanations should be sought for the association between plasma concentrations of beta carotene and diminished risk of death.

Utility of a risk assessment questionnaire in identifying children with lead exposure

OBJECTIVE

To evaluate the utility of the Centers for Disease Control and Prevention (CDC) Risk Questionnaire and a behavioral risk factor questionnaire in identifying children with blood lead concentrations of 0.48 mumol/L (10 micrograms/dL) or more.

DESIGN

Cross-sectional study of 463 urban Massachusetts children (6 to 72 months of age) screened for lead with venous blood.

RESULTS

Twenty-two percent of the children had elevated blood lead concentrations. Of the five CDC questions, only one was significantly associated with an increased adjusted odds ratio for elevated blood lead: having a sibling, housemate, or playmate who was followed up or treated for lead poisoning (odds ratio, 2.7; 95% confidence interval, 1.7 to 4.2; P < .001). Children who had at least one positive or equivocal response to any of the five CDC questions (n = 318 [68.7%]) were not at higher risk than were children who displayed a negative response to all five questions (odds ratio, 1.1; 95% confidence interval, 0.7 to 1.8; P = .69). Of nine behaviors surveyed, two were associated with an increased adjusted odds for elevated blood lead: use of a pacifier (odds ratio, 2.4; 95% confidence interval, 1.3 to 4.4; P = .01) and playing near the outside of the home (odds ratio, 3.4; 95% confidence interval, 2.0 to 5.8; P < .001).

CONCLUSIONS

In this population of children, the CDC risk questionnaire did not identify a group at higher risk for lead exposure. We suggest that practitioners in urban communities screen all children according to the same schedule. We conclude that risk factors differ by community and no risk questionnaire developed at the national level should be applied across communities to target screening.

Effects of intraoperative irradiation and intraoperative hyperthermia on canine sciatic nerve: neurologic and electrophysiologic study

PURPOSE

Late radiation injury to peripheral nerve may be the limiting factor in the clinical application of intraoperative radiation therapy (IORT). The combination of IORT with intraoperative hyperthermia (IOHT) raises specific concerns regarding the effects on certain normal tissues such as peripheral nerve, which might be included in the treatment field. The objective of this study was to compare the effect of IORT alone to the effect of IORT combined with IOHT on peripheral nerve in normal beagle dogs.

METHODS AND MATERIALS

Young adult beagle dogs were randomized into five groups of three to five dogs each to receive IORT doses of 16, 20, 24, 28, or 32 Gy to 5 cm of surgically exposed right sciatic nerve using 6 MeV electrons and six groups of four to five dogs each received IORT doses of 0, 12,16, 20, 24, or 28 Gy simultaneously with 44 degrees C of IOHT for 60 min. IOHT was performed using a water circulating hyperthermia device with a multichannel thermometry system on the surgically exposed sciatic nerve. Neurologic and electrophysiologic examinations were done before and monthly after treatment for 24 months. Electrophysiologic studies included electromyographic (EMG) examinations of motor function, as well as motor nerve conduction velocities studies.

RESULTS

Two years after treatment, the effective dose for 50% complication (ED50) for limb paresis in dogs exposed to IORT only was 22 Gy. The ED50 for paresis in dogs exposed to IORT combined with IOHT was 15 Gy. The thermal enhancement ratio (TER) was 1.5. Electrophysiologic studies showed more prominent changes such as EMG abnormalities, decrease in conduction velocity and amplitude of the action potential, and complete conduction block in dogs that received the combination of IORT and IOHT. The latency to development of peripheral neuropathies was shorter for dogs exposed to the combined treatment.

CONCLUSION

The probability of developing peripheral neuropathies in a large animal model was higher for IORT combined with IOHT, than for IORT alone. The dose required to produce the same level of late radiation injury to the sciatic nerve was reduced by a factor of 1.5 (TER) if IORT was combined with 44 degrees C of IOHT for 60 min.

Long-term morbidity following jejunoileal bypass. The continuing potential need for surgical reversal

OBJECTIVE

To review the late sequelae of jejunoileal bypass (JIB) and the potential role of late surgical reversal in ameliorating morbidity and mortality following JIB.

DESIGN

Patients who underwent JIB between 1965 and 1977 were contacted and pertinent health-event information was gathered. Early sequelae were defined as disorders occurring within the first 2 years after JIB; late sequelae were those occurring after 2 years. Health events occurring between 0 and 23 years after JIB were documented.

SETTING

A private, tertiary referral center.

PATIENTS

Patients underwent JIB for morbid obesity that had failed medical and/or psychiatric interventions.

MAIN OUTCOME MEASURES

Body mass index (BMI) (weight kilograms divided by the square of the height in meters), diarrhea, electrolyte imbalance, acute, and chronic liver disease, renal disease, JIB reversal, reason for JIB reversal, death, and cause of death.

RESULTS

A total of 453 morbidity obese patients underwent JIB. By 2 years following JIB, the mean (+/- SD) BMI dropped from 49.3 +/- 8.1 to 31.1 +/- 0.8 and remained at this level until year 15, after which weight gradually increased (BMI, 35.4 +/- 3.1). The most severe early complication was acute liver failure, which occurred in 7% of patients and caused seven deaths. At 15 years, the actuarial probability of the most common serious late complications related to JIB were renal disease (37%), with two deaths; diarrhea (29%); and liver disease (10%), with three deaths. One hundred thirty-eight patients (31%) had a bypass reversal. The most common indications for reversal were diarrhea and electrolyte disturbance (29%), renal disease (19%), and liver disease (17%). Fifty-six patients died more than 30 days after JIB: 64% before JIB reversal, 13% at the time of reversal, and 23% subsequently.

CONCLUSIONS

Jejunoileal bypass is associated with progressive accrual of serious, sometimes life-threatening complications. Lifelong follow-up for early diagnosis and surgical reversal before life is threatened should reduce the morbidity and mortality associated with this procedure.

Hospital readmission rates for cohorts of Medicare beneficiaries in Boston and New Haven

BACKGROUND

Geographic variations in the use of hospital services are associated with differences in the availability of hospital beds. There continues to be uncertainty about the extent to which unmeasured case-mix differences explain these findings. Previous research showed that the number of occupied beds per capita in Boston was substantially higher than the number of occupied beds per capita in New Haven, Connecticut, and that overall rates of hospital utilization were higher for Boston residents than for New Haven residents.

METHODS

We used Medicare claims data to study cohorts of Medicare beneficiaries 65 years of age or older and residing in Boston or New Haven who were initially hospitalized for one of five indications (acute myocardial infarction, stroke, gastrointestinal bleeding, hip fracture, or potentially curative surgery for breast, colon, or lung cancer). Residents of Boston or New Haven who were discharged between October 1, 1987, and September 30, 1989, were enrolled in the cohort corresponding to the earliest such admission and followed for up to 35 months.

RESULTS

The relative rate of readmission in Boston as compared with New Haven was 1.64 (95 percent confidence interval, 1.53 to 1.76) for all cohorts combined, with a similarly elevated rate for each of the five clinical cohorts and each age, sex, and race subgroup examined. Hospital-specific readmission rates varied substantially among the hospitals in Boston and were higher than those in New Haven. No relation was found between mortality (during the first 30 days after discharge or over the entire study period) and either community or hospital-specific readmission rates.

CONCLUSIONS

Regardless of the initial cause of the admission, Medicare beneficiaries who were initially hospitalized in Boston had consistently higher rates of readmission than did Medicare beneficiaries hospitalized in New Haven. Differences in the severity of illness are unlikely to explain these findings. One possible explanation is a threshold effect of hospital-bed availability on decisions to admit patients.

Standardized rates of recurrent outcomes

Longitudinal studies are often concerned with estimating the rate of an event that may recur. Examples are nonmelanoma skin cancer rates, screening rates for breast cancer using mammography and hospital admission rates. We propose simple estimators for directly and indirectly standardized summary rates and relative rates of recurrent events and their variances. We also develop an estimator of the excess rate in an area if the rate in another area applied. For non-recurrent events, the estimators are identical to the usual standardized summary rates. The estimators are independent of the underlying distribution of the event of interest and allow for unequal follow-up times and event rate heterogeneity among individuals. The method is not computationally intensive and does not require specialized software. We illustrate the application of the method in a retrospective cohort study of hospital utilization patterns of Medicare enrollees in Boston and New Haven over a three and a half year period.

Steady-state serum concentration of alpha tocopherol not altered by supplementation with oral beta carotene. The Polyp Prevention Study 1 Group

BACKGROUND

The antioxidants beta carotene and vitamin E may play a role in cancer prevention. However, some studies have suggested that oral supplements of beta carotene may cause a decrease in serum levels of alpha tocopherol (vitamin E).

PURPOSE

We conducted this study to determine if beta carotene supplements affect serum levels of vitamin E and vice versa.

METHODS

Five hundred five patients in a clinical trial of antioxidant vitamins, used to prevent recurrences of colonic polyps, received either a placebo, 25 mg of beta carotene per day, 1 g ascorbic acid plus 400 mg alpha tocopherol per day, or all three agents combined. Serum levels of beta carotene and vitamin E were measured before and after 9 months of supplementation, using high-performance liquid chromatography.

RESULTS

Vitamin E levels changed very little among the groups receiving placebo or beta carotene and went up substantially and equally in the groups receiving vitamin E plus ascorbic acid or all three agents together. Conversely, beta carotene levels changed very little for the groups receiving placebo or ascorbic acid plus vitamin E but went up substantially and equally for the groups receiving beta carotene alone or all three agents.

CONCLUSIONS

We conclude that oral supplementation with beta carotene for 9 months does not alter serum concentration of vitamin E and that supplementation with vitamin E plus ascorbic acid does not alter serum beta carotene levels.

Normal values for random urinary calcium to creatinine ratios in infancy

OBJECTIVE

To determine normal values for the urinary calcium/creatinine ratio (UCa/Cr) in infants. To assess the impact of short-term supplementation of infant formula with calcium and phosphorus on UCa/Cr in a group of infants.

DESIGN

We determined UCa/Cr in randomly collected urine samples from a group of children and adults. Short-term supplementation of infant formula with calcium glycerophosphate was carried out in 21 infants, and UCa/Cr was monitored in a before-and-after trial.

SETTING

A pediatric clinic at an academic center (infants and adults), and a day-care center (older children).

PARTICIPANTS

A total of 103 infants between 5 days and 7 months of age, 40 infants between 8 and 17 months of age, 41 children between 18 months and 6 years, and 31 adults.

RESULTS

The 95th percentiles for molar UCa/Cr for the different age groups were as follows: less than 7 months, 2.42 (0.86 mg/mg); 7 to 18 months, 1.69 (0.60 mg/mg); 19 months to 6 years, 1.18 (0.42 mg/mg); and adults, 0.61 (0.22 mg/mg). Regression analysis indicated a statistically significant decline in average UCa/Cr with age (R2 = 0.115, p < 0.0001 for log (UCa/Cr) vs log (age)). The geometric means for the two groups of infants were significantly greater than those of the older children and the adults (p < 0.05). Values for UCa/Cr in adults in our sample were comparable to those previously reported. We detected no significant changes in mean UCa/Cr during week-long periods of calcium supplementation of up to 1.8 gm of calcium and 1.39 gm of phosphorus per liter of formula.

CONCLUSION

We conclude that normal values for UCa/Cr are much higher in infants than in older children and adults; UCa/Cr is age-related and declines gradually in the first several years of life, and short-term supplementation of infant formula with calcium glycerophosphate has minimal effect on UCa/Cr.

Comparison of methods for the analysis of longitudinal interval count data

Longitudinal studies are often concerned with estimating the recurrence rate of a non-fatal event. In many cases, only the total number of events occurring during successive time intervals is known. We compared a mixed Poisson-gamma regression method proposed by Thall and a quasi-likelihood method proposed by Zeger and Liang for the analysis of such data, in the case where the mean was correctly specified, using simulation techniques with large samples. Both methods produced similar standard errors in most situations, except in the case of time-dependent covariates with non-Poisson-gamma data where they were seriously underestimated by the Thall method. A simple method for discriminating between the variance forms of the two methods is described. The findings are applied to the analyses of clinical trials of non-melanoma skin cancer and familial polyposis. This study extends the findings of Breslow concerning variance misspecification in overdispersed Poisson and quasi-likelihood models to the longitudinal setting.

A meta-analytic method for summarizing diagnostic test performances: receiver-operating-characteristic-summary point estimates

The authors have devised a meta-analytic method to summarize diagnostic test performances, which they describe along with a clinical example wherein they analyze the performances of real-time ultrasonography in eight studies of the detection of proximal deep venous thrombosis of the lower extremity, selected on the basis of specific inclusion criteria. To evaluate the evidence for fitting a summary receiver operating characteristic (ROC) curve, a test of correlation between the estimates of true-positive rate (TPR) and false-positive rate (FPR) is performed. A high positive correlation argues for summarizing the studies with an ROC curve. In the absence of such correlation, a test of homogeneity is applied separately to the estimates of sensitivity (TPR) and specificity (TNR) to evaluate whether differences among studies are due to chance alone. If the estimates are homogeneous, a summary point estimate and confidence intervals (CIs) are calculated. As a final step, subgroup analyses can be performed to assess alternative explanations of variability in TPR and FPR. Within groups defined by the presence or absence of symptoms of venous thrombosis, a negative correlation between TPR and FPR and homogeneity among studies were found. For symptomatic patients, the summary TPR was 0.97 with a 95% CI of (0.94, 0.99) and the summary TNR was 0.97 (0.95, 0.99). For asymptomatic patients, the summary TPR was 0.66 (0.50, 0.81), and the summary TNR was 0.96 (0.90, 0.99). The difference in TPR between symptomatic and asymptomatic patients was statistically significant. There was no evidence that test referral bias or lack of independent interpretation of test results influenced these findings.(ABSTRACT TRUNCATED AT 250 WORDS)

Estimating the variance of standardized rates of recurrent events, with application to hospitalizations among the elderly in New England

Usual approaches for estimating the variance of a standardized rate may not be applicable to rates of recurrent events. Where individuals are prone to repeated health events, Greenwood and Yule (J R Stat Soc [A], 1920;83:255-79) advocated use of the negative binomial distribution to account for departures from the assumption of randomness of recurrent events required by the Poisson distribution. In this paper, the authors implemented the negative binomial distribution in the computation of annual hospitalization rates within certain hospital market areas. Data used were from 1,549,915 New England residents aged 65 years or more who were enrolled in Medicare between October 1, 1988, and September 30, 1989, and who had 458,593 hospital admissions during that year. New England was partitioned into 170 hospital market areas ranging in population size from 162 to 70,821 elderly Medicare enrollees. The negative binomial distribution demonstrated substantially better fits than the Poisson distribution to the numbers of hospitalizations within hospital market areas. Estimated standard errors for indirectly standardized rates based on the negative binomial distribution were 25-51 percent higher than estimated standard errors that assumed an underlying Poisson distribution. Using regression analysis to smooth overdispersion parameters across hospital market areas produced similar results. The approach described in this paper may be useful in estimation of confidence intervals for standardized rates of recurrent events when these events do not recur randomly.

Immunohistochemical staining for transforming growth factor beta 1 associates with disease progression in human breast cancer

The transforming growth factor beta s (TGF-beta) comprise a family of M(r) 25,000 pluripotent growth factors which have been implicated in the development and progression of human breast cancer. Conflicting data suggest that TGF-beta has the potential to either inhibit or promote the progression of mammary neoplasia. We therefore examined a pathological library of malignant breast biopsy specimens to determine the prevalence and distribution of immunoreactivity with antibodies specific for the three mammalian isoforms of TGF-beta (beta 1, beta 2, and beta 3). We found that intense staining for TGF-beta 1 was positively associated with rate of disease progression, and that this was independent of age, stage, nodal status, or estrogen receptor status (P = 0.009).

Predictive value of flow cytometric analyses of blast cells in assessing the phenotype of the leukemia colony-forming cell (L-CFC) population in acute myeloid leukemia

Acute myeloid leukemia (AML) blast cells (BC) express antigens that are commonly found on their normal counterparts. The leukemia colony-forming cell (L-CFC) subpopulation, identified by its ability to form leukemia colonies in vitro, is thought to be the stem cell population that produces BC. To ascertain the association between myeloid antigens on the BC and the L-CFC from the same patient, we compared the expression of CD14, CD15, CD33, p124 and HLA class I from 17 cases of AML. These particular myeloid antigens were studied because they are suitable targets in purging bone marrow for autotransplantation. We found no significant difference in the expression of CD14, CD15, CD33, and HLA class I on the BC and L-CFC from the same patient, although we observed considerable heterogeneity among different AML cases. Analysis of the progenitor cell antigen p124 revealed significant within-patient differences on the BC and L-CFC (p = 0.007), with a greater tendency for expression on the L-CFC. This heterogeneity may be due to differences in maturation stage of the L-CFC and BC. This information is important when L-CFC phenotype is used to determine the appropriate selection of antibodies for purging of residual disease in the context of auto-transplantation.

Risk of subsequent basal cell carcinoma and squamous cell carcinoma of the skin among patients with prior skin cancer. Skin Cancer Prevention Study Group

OBJECTIVE

The primary aims of this study were to assess risk of subsequent basal and squamous cell skin cancer among patients with a prior history of these tumors and to examine these risks in relation to patient characteristics and life-style factors.

DESIGN

Follow-up of participants in a randomized trial of betacarotene as a possible skin cancer preventive agent.

SETTING

Clinical centers in Los Angeles, Calif, San Francisco, Calif, Minneapolis, Minn, and Hanover, NH.

PARTICIPANTS

Patients (n = 1805) who were diagnosed as having a basal or squamous cell skin cancer between January 1980 and February 1986 and were free of skin cancer at study entry.

MAIN OUTCOME MEASURE

Time from study entry to first new occurrence of basal and squamous cell skin cancer.

RESULTS

The estimated risk of developing one or more new skin cancers was 35% at 3 years and 50% at 5 years. New skin cancers tended to be of the same cell type as the previous skin cancers. For both basal and squamous cell skin cancer, risk was higher among patients who were male, were over the age of 60 years, had more prior skin cancers, had severe actinic skin damage, or who burned easily with sun exposure. Compared with those who had never smoked, the rate of subsequent squamous cell skin cancer was higher among current smokers (rate ratio, 2.01; 95% confidence interval, 1.21 to 3.34) and former smokers (rate ratio, 1.62; 95% confidence interval, 1.07 to 2.47) and increased with both duration and amount smoked. There was no clear relationship between smoking and basal cell skin cancer; the rate appeared lower among heavy smokers but was unrelated to duration of smoking.

CONCLUSIONS

Persons with a prior nonmelanoma skin cancer had a substantial 5-year risk of developing another tumor of the same histologic type. Number of previous skin cancers, solar damage, and skin sensitivity to sun exposure were particularly related to this risk. The increased risk of squamous cell skin cancer associated with cigarette smoking merits further study.

Diurnal and seasonal variation of five carotenoids measured in human serum

We studied within-person variation over time in serum concentrations of five carotenoids. In a diurnal study involving 33 subjects, only the 1700 h blood samples demonstrated carotenoid concentrations different from the original 0800 values. Correlations between serum concentrations of the same carotenoids drawn 1 d apart ranged from 0.93 to 0.98. In a seasonal study involving 29 subjects, no systematic trends were observed for serum concentrations of these carotenoids. Correlations between concentrations of the same carotenoids drawn 1 y apart ranged from 0.57 to 0.82. Concentrations of different carotenoids within an individual tended to be correlated with each other. Obtaining one blood sample from subjects is a relatively imprecise way to estimate their usual serum concentrations of carotenoids. If an epidemiological study was to be based on only one determination of serum carotenoids, within-person variability in serum concentrations would attenuate true regression coefficients by 4-13% and would increase the required numbers of study subjects by 19-65%.