Continuous versus group sequential analysis for post-market drug and vaccine safety surveillance.

The use of sequential statistical analysis for post-market drug safety surveillance is quickly emerging. Both continuous and group sequential analysis have been used, but consensus is lacking as to when to use which approach. We compare the statistical performance of continuous and group sequential analysis in terms of type I error probability; statistical power; expected time to signal when the null hypothesis is rejected; and the sample size required to end surveillance without rejecting the null. We present a mathematical proposition to show that for any group sequential design there always exists a continuous sequential design that is uniformly better. As a consequence, it is shown that more frequent testing is always better. Additionally, for a Poisson based probability model and a flat rejection boundary in terms of the log likelihood ratio, we compare the performance of various continuous and group sequential designs. Using exact calculations, we found that, for the parameter settings used, there is always a continuous design with shorter expected time to signal than the best group design. The two key conclusions from this article are (i) that any post-market safety surveillance system should attempt to obtain data as frequently as possible, and (ii) that sequential testing should always be performed when new data arrives without deliberately waiting for additional data.

Automated use of WHONET and SaTScan to detect outbreaks of Shigella spp. using antimicrobial resistance phenotypes.

Antimicrobial resistance is a priority emerging public health threat, and the ability to detect promptly outbreaks caused by resistant pathogens is critical for resistance containment and disease control efforts. We describe and evaluate the use of an electronic laboratory data system (WHONET) and a space-time permutation scan statistic for semi-automated disease outbreak detection. In collaboration with WHONET-Argentina, the national network for surveillance of antimicrobial resistance, we applied the system to the detection of local and regional outbreaks of Shigella spp. We searched for clusters on the basis of genus, species, and resistance phenotype and identified 19 statistical 'events' in a 12-month period. Of the six known outbreaks reported to the Ministry of Health, four had good or suggestive agreement with SaTScan-detected events. The most discriminating analyses were those involving resistance phenotypes. Electronic laboratory-based disease surveillance incorporating statistical cluster detection methods can enhance infectious disease outbreak detection and response.

Power Evaluation of Focused Cluster Tests.

Many statistical tests have been developed to assess the significance of clusters of disease located around known sources of environmental contaminants, also known as focused disease clusters. The majority of focused-cluster tests were designed to detect a particular spatial pattern of clustering, one in which the disease cluster centers around the pollution source and declines in a radial fashion with distance. However, other spatial patterns of environmentally related disease clusters are likely given that the spatial dispersion patterns of environmental contaminants, and thus human exposure, depend on a number of factors (i.e., meteorology and topography). For this study, data were simulated with five different spatial patterns of disease clusters, reflecting potential pollutant dispersion scenarios: 1) a radial effect decreasing with increasing distance, 2) a radial effect with a defined peak and decreasing with distance, 3) a simple angular effect, 4) an angular effect decreasing with increasing distance and 5) an angular effect with a defined peak and decreasing with distance. The power to detect each type of spatially distributed disease cluster was evaluated using Stone's Maximum Likelihood Ratio Test, Tango's Focused Test, Bithell's Linear Risk Score Test, and variations of the Lawson-Waller Score Test. Study findings underscore the importance of considering environmental contaminant dispersion patterns, particularly directional effects, with respect to focused-cluster test selection in cluster investigations. The effect of extra variation in risk also is considered, although its effect is not substantial in terms of the power of tests.

Using encounters versus episodes in syndromic surveillance.

BACKGROUND: Automated electronic medical records may be useful for syndromic surveillance to quickly detect infectious disease outbreaks. Some syndromic surveillance systems include every encounter in the analysis, whereas others exclude individuals' repeat encounters within the same syndrome occurring within a short period of time, with the rationale that these represent follow-up visits rather than new episodes of illness. METHODS: We evaluate the effect of keeping all encounters as compared with removing repeat encounters. Using the prospective space-time permutation scan statistic, we performed daily analyses on all encounters versus on episodes defined as encounters new within 2, 6 or 12 weeks. Data were taken from a Massachusetts Health Maintenance Organization (HMO) for the calendar year 1999 for four different syndromes. RESULTS: We found extensive disagreement in the number of signals detected: 70, 68, 21 and 15 signals when using all encounters versus 15-20, 3, 4-5 and 0 signals when using only new episodes for lower respiratory, lower gastrointestinal, upper gastrointestinal and neurologic syndromes, respectively. CONCLUSION: Using all encounters in syndromic surveillance may not only create too many signals but may also miss some signals by masking the anomalies generated by actual episodes. However, it is also possible to miss signals when using episodes.

Ambulatory-care diagnoses as potential indicators of outbreaks of gastrointestinal illness--Minnesota.

INTRODUCTION: Syndromic surveillance's capability to augment existing surveillance for community-acquired gastrointestinal disease is unknown. OBJECTIVE: The objective of this study was to evaluate the capability of a syndromic surveillance system to detect outbreaks of gastrointestinal disease. METHODS: A retrospective analysis was conducted comparing ambulatory care data from a health plan with a set of 110 gastrointestinal-disease outbreaks identified by the Minnesota Department of Health during 2001-02. Unusual clusters of illness (i.e., signals) in the health-plan data were identified by analyzing daily counts of gastrointestinal illness using an adjusted space-time scan statistic. Concordance was defined as < or =5 km between outbreak and signal and the signal occurring within 1 week of the outbreak. RESULTS: During 104 weeks, the number of signals was roughly what would have been expected by chance, suggesting that the modeling did a good job of estimating the expected counts of illness and that false alarms would not have occurred much more often than the number predicted at the various thresholds. During the same period, the health department identified 110 eligible gastrointestinal outbreaks. Apparent associations of the three statistically most unusual concordant signals with outbreaks of viral or bacterial gastrointestinal illness were ruled out by the health department on the basis of detailed knowledge of the circumstances and low numbers of affected persons seeking medical care. CONCLUSION: No previously known gastrointestinal outbreaks were identified by this surveillance system. However, relatively few recognized outbreaks resulted in patients seeking medical care, and the sensitivity of this system to detect outbreaks of real significance to public health remains to be determined. Prospective evaluation probably will be required to understand the usefulness of syndromic surveillance systems to enhance existing disease surveillance.

A model-adjusted space-time scan statistic with an application to syndromic surveillance.

The space-time scan statistic is often used to identify incident disease clusters. We introduce a method to adjust for naturally occurring temporal trends or geographical patterns in illness. The space-time scan statistic was applied to reports of lower respiratory complaints in a large group practice. We compared its performance with unadjusted populations from: (1) the census, (2) group-practice membership counts, and on adjustments incorporating (3) day of week, month, and holidays; and (4) additionally, local history of illness. Using a nominal false detection rate of 5%, incident clusters during 1 year were identified on 26, 22, 4 and 2% of days for the four populations respectively. We show that it is important to account for naturally occurring temporal and geographic trends when using the space-time scan statistic for surveillance. The large number of days with clusters renders the census and membership approaches impractical for public health surveillance. The proposed adjustment allows practical surveillance.

Dietary intake of heterocyclic amines, meat-derived mutagenic activity, and risk of colorectal adenomas.

Meats cooked well-done by high temperature techniques produce mutagenic compounds such as heterocyclic amines (HCAs), but the amounts of these compounds vary by cooking techniques, temperature, time, and type of meat. We investigated the role of HCAs in the etiology of colorectal adenomas and the extent to which they may explain the previously observed risk for red meat and meat-cooking methods. In a case-control study of colorectal adenomas, cases (n = 146) were diagnosed with colorectal adenomas at sigmoidoscopy or colonoscopy, and controls (n = 228) were found not to have colorectal adenomas at sigmoidoscopy. Using a meat-derived HCA and mutagen database and responses from a meat-cooking questionnaire module, we estimated intake of 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (DiMeIQx), 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and mutagenic activity. We calculated odds ratios and 95% confidence intervals using logistic regression adjusting for several established risk factors for colorectal adenomas or cancer. The odds ratios (95% confidence interval; P for trend test) fifth versus first quintiles are: 2.2 (1.2-4.1; P = 0.02) for DiMeIQx; 2.1 (1.0-4.3; P = 0.002) for MeIQx; 2.5(1.1-5.5; P = 0.02) for PhIP; and 3.1 (1.4-6.8; P = 0.001) for mutagenic activity. When the three HCAs were adjusted for the other two, only the trend for MeIQx (P = 0.04) remained statistically significant. When we tried to disentangle the relative contribution of the three HCAs from the meat variables, we found that MeIQx remained significantly associated with risk even when adjusted for red meat but not vice versa. When MeIQx and well-done meat were analyzed in the same model, the risks were attenuated for both. Mutagenic activity from meat remained significantly associated with increased risk even when adjusted for intake of red meat or well-done red meat, whereas the red meat and well-done red meat associations were no longer significant when adjusted for total mutagenic activity. In conclusion, we found an elevated risk of colorectal adenomas associated with high intake of certain HCAS: Further, mutagenic activity from cooked meat consumption, a measure that integrates all of the classes of mutagens, was strongly associated with risk and explained the excess risk with intake of well-done red meat.

Geographical differences in primary therapy for early-stage breast cancer.

BACKGROUND: Breast-conserving surgery may not be uniformly available to all women. We evaluated geographical differences across Connecticut in the proportions of cases with early stage breast cancer treated by partial mastectomy (PM). We also looked at geographical variation in PM with axillary lymph node dissection (AND) and PM with adjuvant radiotherapy (RAD). METHODS: The Connecticut Tumor Registry identified 9106 cases of early disease for 1991 to 1995. Latitude-longitude coordinates for place of residence at diagnosis and initial form of therapy were available for 8795 records. A spatial scan statistic was used to detect geographical differences in treatment rates across the state. RESULTS: A total of 57.7% of early breast cancer cases were treated by PM. Women living around New Haven were less likely than others to be treated in that manner (relative risk [RR] = .86; P = .0001), whereas those living around Norwalk were more likely (RR = 1.26; P = .0001). PM with AND, relative to PM alone, was reported less often for cases from a large area of southwestern Connecticut (RR = .89; P = .0001), but more often for those in north central Connecticut (RR = 1.13; P = .0001). PM with RAD, relative to PM alone, was less common for cases around Danbury (RR = .40; P = .0001) but more common among cases around Hartford (RR = 1.14; P = .0001). CONCLUSIONS: Geographical analysis is a way for physicians and health officials to identify groups of women who may not yet benefit from preferred surgical procedures.

Comparing odds ratios for nested subsets of dietary components.

BACKGROUND: In nutritional epidemiology, it is often of interest to disentangle the risk of disease associated with related foods or nutrients, where the food items are in a nested arrangement within a larger group. To compare odds ratios (OR) derived from a standard quantile-based analysis can be misleading since the amounts consumed may differ substantially for different dietary components. METHODS: The authors applied different logistic regression models on a case-control study concerning the risk of colorectal adenomas due to meat and its different subsets such as white meat, red meat and well-done red meat. RESULTS: By calculating OR for a fixed amount of intake, the authors suggest a method for partitioning the risk of one dietary item into that associated with increasingly detailed sub-components. A graph is presented for illustrating such partitions in terms of both addition and substitution effects. CONCLUSIONS: Odds ratios based on upper versus lower quantiles or percentiles are useful as they compare the risk between the upper and lower ends of the consumption range. A complimentary set of OR are those based on fixed amounts of consumption. These allow for direct comparisons between nested subgroups of dietary components, in order to disentangle the risk linked to specific groups of foods or nutrients.

Geographic variation in prostate cancer mortality rates among white males in the united states.

PURPOSE: The most recent atlas of cancer mortality in the United States revealed elevated prostate cancer mortality rates among white males in the northwest, Rocky Mountain, northcentral, and southeast areas, as well as New England, especially during the 1970-94 period. We wanted to test whether this observed geographic variation was simply due to chance or not.METHODS: We used a spatial scan statistic using mortality data for 506 state economic areas.RESULTS: There were four significant clusters with elevated risks of prostate cancer mortality (P < 0.001). The most prominent cluster was in the northwestern quadrant of the country, followed by clusters in New England, the midwest, and southeast regions. Within the northwestern cluster, we also detected seven significant sub-clusters (P < 0.05).CONCLUSIONS: We concluded that the observed geographic variation of prostate cancer mortality is indeed real, and deserves further study into the underlying determinants.

Measuring cell proliferation in the rectal mucosa. comparing bromodeoxyuridine (BrdU) and proliferating cell nuclear antigen (PCNA) assays.

Cell proliferation in the human colorectum can be measured using bromodeoxyuridine (BrdU) or proliferating cell nuclear antigen (PCNA) assays. Using data from the National Cancer Institute's Polyp Prevention Trial, these two assays are compared using correlation coefficients and variance components analysis. Adjusting for fixed as well as for the random effects of between-biopsy and scoring variation, the estimated correlation is 0.46 for the log labeling index and 0.45 for log proliferative height. This is an estimate of the highest correlation that can be achieved by taking multiple biopsies scored by multiple scorers. For single biopsies, the estimated correlation is 0.16 and 0.10, respectively. There are significant differences between the variance components for the two assays. For example, for log labeling index, PCNA has a lower variation between biopsies than BrdU, but higher variation between scorings. When used in a clinical or epidemiological setting, it is important to take multiple biopsies at multiple time points.

Evaluation of alternative approaches to assign nutrient values to food groups in food frequency questionnaires.

Although every food frequency questionnaire (FFQ) requires a nutrient database to produce nutrient intake estimates, it is often unclear how a particular database has been generated. Moreover, alternative methods for constructing a database have not been rigorously evaluated. Using 24-hour recalls from the 1994-1996 Continuing Survey of Food Intake by Individuals, the authors categorized 5,261 individual foods reported by 10,019 adults into 170 food groups consistent with line items on an FFQ. These food groups were used to generate 10 potential nutrient databases for a FFQ that varied by whether the authors 1) used means or medians, 2) did or did not consider age, 3) incorporated collapsing strategies for small age-gender-portion size cells, 4) excluded outliers in a regression, and 5) used weighted median nutrient density x age-gender-portion size-specific median gram weights (Block method). Mean error, mean squared error, and mean absolute error were calculated and compared across methods, with error being the difference in total observed (from recalls for each individual) and total estimated intake (from each of the 10 methods) for seven nutrients. Mean methods for assigning nutrients to food groups were superior to median approaches for all measurements. Among the mean methods, no single variation was consistently better.

Evaluation of spatial filters to create smoothed maps of health data.

Spatial filters have been used as an easy and intuitive way to create smoothed disease maps. Birth weight data from New York State for 1994 and 1995 are used to compare the traditional filter type of fixed geographical size with a filter size of constant or nearly constant population size. The latter are more appropriate for mapping disease in geographic areas with widely varying population density, such as New York State. Issues such as the choice of population size for the filter, the scale of smoothing, the ability to detect true spatial variation and the ability to smooth over random spatial noise are evaluated and discussed.

Dietary heterocyclic amines and the risk of lung cancer among Missouri women.

Heterocyclic amines (HCAs) such as 2-amino-3,8-dimethylimidazo[4,5f]quinoxaline (MeIQx), 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (DiMeIQx,), and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) are found in meats cooked at high temperatures. In rodents, MeIQx induces lung tumors. The purpose of this study was to investigate lung cancer risk posed by different HCAs in the diet. A population-based case-control study of 593 cases and 623 frequency-matched controls including both nonsmoking and smoking women was conducted in Missouri. An administered food frequency questionnaire with detailed questions on meat consumption, degrees of internal doneness, surface browning/charring, and cooking technique was linked to a database that provided exposure estimates of three HCAs. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using logistic regression. When comparing the 90th and 10th percentiles, significant excess risks were observed for MeIQx (OR, 1.5; CI, 1.1-2.0), but not for DiMeIQx (OR, 1.2; CI, 0.9-1.6) or PhIP (OR, 0.9; CI, 0.8-1.1). MeIQx consumption was associated with increased risk of lung cancer for nonsmokers (OR, 3.6; CI, 1.3-10.3) and light/moderate smokers (OR, 2.1; CI, 1.3-3.3), but not for heavy smokers (OR, 1.0; CI, 0.7-1.5). There was elevated risk with MeIQx intake for subjects with squamous cell carcinomas (OR, 1.9; CI, 1.2-3.1) and "other histological cell types" (OR, 1.6; CI, 1.1-2.5), but not for subjects with small cell carcinomas and adenocarcinomas. Neither DiMeIQx nor PhIP showed an association with smoking categories or lung cancer histology. In conclusion, MeIQx may be associated with lung cancer risk, but DiMeIQx and PhIP are probably not associated with lung cancer risk.

Dietary fat, fat subtypes, and breast cancer in postmenopausal women: a prospective cohort study.

BACKGROUND: The intake of total dietary fat and of certain fat subtypes has been shown to be strongly associated with breast cancer in international comparisons and in animal experiments. However, observational epidemiologic studies have generally reported either weak positive or no associations. To extend the prospective epidemiologic evidence on this question, we examined the association between adult dietary intake of fat, fat subtypes, and breast cancer in a large, prospective cohort of postmenopausal women. METHODS: Participants were selected from a national breast cancer mammography screening program conducted from 1973 through 1981 at 29 centers throughout the United States. From 1987 through 1989, 40022 postmenopausal women satisfactorily completed a mailed, self-administered questionnaire that included a 60-item National Cancer Institute/Block food-frequency questionnaire. Women were then followed for an average of 5.3 years; 996 women developed breast cancer. Risk was assessed by use of Cox proportional hazard regression, with age as the underlying time metric. All statistical tests were two-sided. RESULTS: Compared with women in the lowest quintile (Q1) of percentage of energy from total fat, the adjusted risk ratio (RR) and 95% confidence interval (CI) for women in the highest quintile (Q5) was 1.07 (95% CI = 0.86-1.32). In analyses stratified by history of benign breast disease (BBD), a positive association was observed among only women with no history of BBD (RR (Q5 versus Q1) = 2.20; 95% CI = 1.41-3.42; test for trend, P =.0003). The increased risk in these women appeared to be attributable to unsaturated fat intake and oleic acid in particular. CONCLUSIONS: In this study, there was no overall association between fat intake during adulthood and breast cancer risk; however, among women with no history of BBD, there appeared to be a positive association between total and unsaturated fat intake and breast cancer risk.

The effect of low dose micronized 17ss-estradiol on bone turnover, sex hormone levels, and side effects in older women: a randomized, double blind, placebo-controlled study.

The purpose of this study was to examine the effects of three doses (0.25, 0.5, and 1.0 mg/day) of micronized 17ss-estradiol on bone turnover, sex hormone levels, and side effects compared with placebo in healthy older women. The study design was randomized, double blind, and placebo controlled. The setting was a university clinical research center. Healthy, community-living women over 65 yr of age participated in the study. The main outcome measures were serum and urinary biochemical markers of bone resorption and formation at baseline, 6 and 12 weeks on treatment, and 6 and 12 weeks off treatment. Markers of bone resorption were N-telopeptides of type I collagen, C-telopeptides of type I collagen, and total deoxypyridinoline cross-links; formation markers were bone alkaline phosphatase, osteocalcin, and N-terminal procollagen peptides. We also measured serum estradiol, estrone, and sex hormone-binding globulin levels at baseline, 12 weeks on treatment, and 12 weeks posttreatment. All markers of bone resorption significantly decreased at 12 weeks on treatment compared with placebo and returned toward baseline at 12 weeks posttreatment. Two markers of bone formation, bone alkaline phosphatase and N-terminal procollagen peptides, significantly decreased 12 weeks posttreatment, but the decrease in osteocalcin varied with time and estrogen dose. Based on equivalence testing, the response of markers of bone turnover to therapy with 0.25 mg/day was similar to that seen with 1.0 mg/day. Serum estradiol increased compared with baseline in all treatment groups and compared with placebo in the two higher dose groups. Breast tenderness, bleeding, and endometrial changes were significantly less frequent in the 0.25 mg/day and placebo groups compared with the higher dose groups. We conclude that low dose of estrogen (0.25 mg/day 17ss-estradiol) reduced bone turnover to a similar degree as that seen with usual replacement therapy (1.0 mg/day 17ss-estradiol), but had a side effect profile similar to that of placebo. In our study additional increases in estradiol levels, as seen with 0.5 and 1.0 mg/day 17ss-estradiol treatment, resulted in more side effects without evidence of additional benefit to bone. These data suggest that 0.25 mg/day 17ss-estradiol may be an effective and tolerable agent for the treatment of osteoporosis in older women.

Geographic assessment of breast cancer screening by towns, zip codes, and census tracts.

Early detection is the primary way to control breast cancer, and mammography screening can reduce breast cancer mortality 30 to 40 percent among women aged 50 years and older. Geographic areas with a high proportion of cases with late-stage diagnoses may reflect gaps in screening efforts. We used a spatial scan statistic, adjusting for the multitude of possible region locations and sizes, to test whether any particular region of Massachusetts had statistically significant excesses of late-stage diagnoses during the period 1982 to 1986. The novel geographic analysis technique utilized here can also be used in the control of other types of cancer.

Maternal supplemental and dietary zinc intake and the occurrence of neural tube defects in California.

The authors investigated the association between maternal preconceptional supplemental and dietary zinc intake and risk of neural tube defects (NTDs) in a population-based case-control study conducted between 1989 and 1991 in California. Cases were 430 NTD-affected fetuses/infants, and controls were 429 randomly selected non-malformed infants. Mothers reported their preconceptional use of vitamin, mineral, and food supplements, and completed a 98-item food frequency questionnaire. Increased total preconceptional zinc intake was associated with a reduced risk for NTDs (quintile 5 vs. quintile 1, odds ratio (OR) = 0.65, 95% confidence interval (CI) 0.43, 0.99). Phytate intake, a constituent of the diet known to impede zinc absorption, appeared to modify the zinc - NTD association. In addition, increased servings of animal products, the most bioavailable food source of zinc, was associated with a reduced risk for NTDs (quintile 5 vs. quintile 1, OR = 0.49, 95% CI 0.32, 0.76). Risk estimates for zinc intake were changed little after controlling for multiple sociodemographic factors and total folate intake, but were attenuated after controlling for nutrients highly correlated with dietary sources of zinc, such as protein. In sum, the analyses indicate that risk of NTDs in infants and fetuses decreased with increasing maternal preconceptional zinc intake. However, it remains unclear whether increased zinc intake, or another nutrient or combination of nutrients highly correlated with zinc intake in the diet, is causally associated with reduced NTD risk.