Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Statistical Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,783 records · Page 99Linked to original sources
PubMed1985

Prediction of staphylococcal etiology among patients with septicemia with or without endocarditis by multivariate statistical methods.

Multivariate statistical methods, multiple regression (RA) and automatic interaction detector analysis (AID) were used to study the possibility of an early prediction of staphylococcal etiology in 249 of 851 patients with verified septicemia or endocarditis. The variables included pertinent symptoms and signs and laboratory data available soon after admission. 10 of the 70 variables initially studied showed simple, or in various combinations, a statistically significant partial correlation to staphylococcal etiology in the AID. The highest predictive value with a high probability for staphylococcal etiology was recorded for combinations of the variables: i.v. narcotic addiction and septic pulmonary embolism; non-addiction, wound infection, and hospitalization within 4 weeks; non-addiction, absence of skin infection, presence of foreign body, and age less than 60 yr. Staphylococcal etiology was contradicted by the absence of i.v. narcotic addiction, skin infection, foreign body, septic skin manifestation, surgical procedure within 4 weeks, joint symptom and a C-reactive protein less than or equal to 10 mm. Thus, a prediction of etiology may be valuable in choosing therapy before definite confirmation by positive blood cultures or when blood cultures remain sterile.

Adolescent↗
PubMedDate not supplied

The map comparison problem: tests for the overlap of geographic boundaries.

The quality of environmental studies is often compromised by the use of statistics, such as correlation and regression for example, which presuppose a statistical model, linear or otherwise, between two variables. When investigating hypotheses about relationships among geographically distributed variables, an alternative approach is to measure the amount of boundary overlap. Boundaries are geographic zones of rapid change in the intensity of a variable, and are often of scientific interest in their own right. Examples of boundaries include ecotones, genetic hybrid zones, pollution plumes, and the front of the wave of advance of an epidemic. Boundary overlap describes zones where boundaries from two or more variables coincide, and are useful for evaluating epidemiologic hypotheses relating health to environmental exposures. This paper proposes four statistics of boundary overlap, and explores their performance using simulation models and real data describing ozone concentrations and hospital admissions for respiratory conditions. The statistics are found sensitive to different aspects of boundary overlap, and provide an additional diagnostic tool in the analysis of geographically distributed variables. Overlap statistics are expected to come into increasing use as the installed base of geographic information systems increases.

Air Pollution↗
PubMed2001

Benefits-based analysis of visitor use of Sorak-san National Park in Korea.

Benefits-based management seeks to provide recreation benefits for recreation participants by managing the physical environments in which recreation occurs. This study investigates the relationship between benefits desired by visitors and the physical, social, and managerial characteristics of settings that facilitate realization of recreation benefits. Data such as perceived benefits from recreation experiences, setting attributes that significantly influenced perceived benefits, and sociodemographic variables were collected from 376 visitors to Sorak-san National Park in the eastern part of the Korean peninsula. Cluster analysis was used to group visitors' desired benefits into 12 domains: relationship with nature/scenery, escaping pressure, learning about nature, family togetherness, introspection, exploration, autonomy/achievement, being with friends, leading others, skills/learning, risk taking, and meeting/observing new people. Multiple regression was used to link benefit domains with the characteristics of settings. The social attribute of recreation settings was linked to eight of the ten benefit domains. There were two statistically strong multiple regression correlations: (1) between domain of "relationship with nature/scenery" and the attributes "forest/water," "attractive nature," and "facility/maintenance" and (2) between the domain of "escaping pressure" and the attributes "attractive nature" and "social." The results of this study are useful to managers in their efforts to provide recreation opportunities for visitors to achieve beneficial outcomes.

Adolescent↗
PubMed1999

Neonatal outcome of inborn and transported very-low-birth-weight infants: relevance of perinatal factors.

OBJECTIVE: To compare the neonatal outcome (survival, intraventricular hemorrhage and bronchopulmonary dysplasia) of inborn and outborn very-low-birth-weight infants accounting for sociodemographic, obstetric and perinatal variables. STUDY DESIGN: Ninety-one premature infants with birth weights of 750-1250 g delivered between 1990 and 1994 in a hospital providing neonatal intensive care were compared with 76 premature babies delivered in a referring hospital. In the statistical analysis, variables with a statistically significant association with the outcome variables and dissimilar distributions in the two hospitals were identified and entered together with the hospital of birth as explanatory variables in a logistic regression. RESULTS: No statistically significant differences between the outcome variables of the two populations examined were observed, whether before or after accounting for the covariates. The odds ratios (outborns relative to inborns) were 1.18 for mortality, 1.25 for bronchopulmonary dysplasia and 1.53 for severe intraventricular hemorrhage. In the multivariate analyses, respiratory distress syndrome was significantly associated with mortality; both low birth weight and the presence of respiratory distress syndrome were associated with the development of bronchopulmonary dysplasia; the evolvement of severe intraventricular hemorrhage was associated with respiratory distress syndrome, initial low Apgar score, advanced multiparity and delivery at the 28-29th week compared to the 23rd-27th week. Antenatal steroid administration had a protective effect. CONCLUSION: Our results concur with the notion that a tertiary center is the optimal location for delivery of the high risk neonate. Improvement in medical and nursing care prenatally and at delivery and transportation, including frequent administration of antenatal steroids and earlier administration of surfactant prior to transportation, may minimize the disadvantage of delivery in a referring hospital.

Hospitals, General↗
PubMed2000

Reactive bone marrow changes in infectious spondylitis: quantitative assessment with MR imaging.

PURPOSE: To evaluate diffuse, reactive bone marrow changes in unaffected vertebrae on magnetic resonance (MR) images in patients with proved infectious spondylitis. MATERIALS AND METHODS: Percentage signal intensity increase of the unaffected bone marrow on contrast material-enhanced MR images (percentage enhancement) was calculated retrospectively in 22 cases of infectious spondylitis and 86 cases without bone marrow disease. Multiple regression analysis and Student t test statistics were performed. RESULTS: Multiple regression analysis showed a significant influence of age and the presence of spondylitis on the values of percentage enhancement (P: <.001). For those aged 35 years or younger, the mean percentage enhancement was 43.2% +/- 4.0 for patients with infectious spondylitis (n = 3) and was 26.4% +/- 8.6 for the control group (n = 23). For those older than 35 years, the mean percentage enhancement was 28.2% +/- 12.2 for patients with infectious spondylitis (n = 19) and 17.5% +/- 7.9 (P: <.001) for the control group (n = 63). Six (27%) of 22 patients with infectious spondylitis showed abnormal percentage enhancement values in unaffected bone marrow when the upper limit of the normal value was 2 SDs above the mean of the control group. CONCLUSION: On MR images, reactive bone marrow changes can be found in unaffected vertebrae in patients with infectious spondylitis. The signal intensity changes and increased percentage enhancement associated with this disease are similar to those of myeloproliferative and diffuse neoplastic disorders and bone marrow stimulation in hemolytic anemia.

Adult↗
PubMedDate not supplied

Erythropoietin requirements: a comparative multicenter study between peritoneal dialysis and hemodialysis.

BACKGROUND: The management of anemia with erythropoietin (EPO) is important in the global treatment of dialysis patients. There is a general impression that anemia control with EPO is obtained more easily in peritoneal dialysis (PD) patients than in hemodialysis (HD) patients. The EPO administration route has to be the same to compare the two techniques adequately. METHODS: To compare EPO action by subcutaneous (SC) route in HD and PD, 132 stable patients were recruited (HD: 69, PD: 63) from six centers, with adequate dialysis criteria (Kt/V in HD >1.3; weekly Kt/V in PD >1.8). In a cross-sectional study, the EPO dose/week, the number of EPO doses/week, hemoglobin (Hb), ferritin, transferrin saturation index (TS), albumin and intact parathyroid hormone (iPTH) were analyzed. Iron treatment, comorbidity and ACE inhibitors (ACEI) and angiotensin II antagonist (AIIA) treatment were recorded. A multivariate regression model was used in the statistical analysis. RESULTS: The mean Hb level was the same in both groups, HD 11.6 (1.3) g/dL, PD 11.4 (1.4) g/dL, p=0.3. The SC, EPO doses required to obtain the Hb levels were higher in HD than in PD patients, with a difference of 64.3 u/Kg/week, statistically significant in the multivariate regression model (p=0.001, 95% CI 42.6-86.0). The number of EPO doses/week was also higher in HD patients (65% of HD patients with > or = 3 doses, 19% of PD patients with three or more doses, p<0.001). TS was similar in both groups, while ferritin was higher in HD patients, with a higher percentage of HD patients using intravenous (i.v.) iron (HD 77% vs. PD 49%, p=0.001). Serum albumin and iPTH were lower in PD patients (p<0.001 and p=0.04, respectively), but the percentage of patients with intact parathyroid hormone (iPTH) >500 pg/mL was similar in both groups (HD 17%, PD 14%). CONCLUSIONS: With the same administration route, PD patients showed a reduced EPO requirement, and less frequent EPO administration than HD patients, to obtain the same Hb level. No other factors, except those involved in better depuration of erythropoiesis inhibitors in PD, seemed responsible for the different EPO requirements.

Aged↗
PubMed1996

Do routinely registered preoperative data provide prognostic information on the short-term outcome of distal bypass surgery?

One hundred and thirty and patients were operated on with unilateral distal bypass procedures for critical ischemia. Fifty-one patients (39%) ended up with graft occlusion or required major amputation within six months after reconstruction. Two-year graft patency was 38%. These rather disappointing results may be partly ascribed to inadequate preoperative selection of patients. Logistic regression model was used to assess whether the independent variables age, gender, diabetes mellitus, smoking, ankle-brachial pressure index (ABI), type of graft, coronary heart disease (CHD), previous vascular operation and site of distal anastomosis were associated with the outcome amputation and graft occlusion six months after reconstruction. CHD, type of graft, ABI and the location of distal anastomosis were significantly associated with outcome. The odds for amputation or occlusion was 2.4 times higher in the CHD group, 4 times higher if the anastomosis was located to the peroneal artery and 2.6 times higher for synthetic grafts. The logistic regression model was statistical significant (p = 0.03). However, the model did not aid sufficiently in the prediction of outcome, since one third was erroneously classified as success or failure. Cox's proportional hazard regression was employed to estimate the influence of the independent variables on graft patency. Favorable patency was found for those with distal anastomosis located to the tibial arteries, the non-CHD group and for higher ABI values. There was a trend towards significance for better patency in the autogenous vein group (p = 0.06). Although combinations of risk factors usable for preoperative prediction of the likelihood of graft failure, appropriate selection models of patients suitable for distal bypass operation have to be improved, to minimize the number of failed procedures.

Aged↗
PubMed1993

Anterior chamber depth and lens thickness in primary angle-closure glaucoma: a case-control study.

Anterior chamber depth and lens thickness have been considered as important biometric determinants in primary angle-closure glaucoma (PACG). In a tertiary care centre-based case-control study, 70 patients and equal number of controls were investigated to analyse the strength of association and predictability of anterior chamber depth (ACD) and lens thickness (LT) in the disease. Mean (+/- S.D.) ACD and LT in the cases and the controls were found to be 2.28 +/- 0.19, 2.87 +/- 0.10; 4.57 +/- 0.34 and 4.13 +/- 0.19 mm respectively. Two sample t test demonstrated statistically significant difference in the ACD and LT between the cases and the controls (Difference being -0.59, 0.44; 95% confidence interval of the difference: -0.64, -0.53 and 0.34, 0.53 respectively, P < 0.01). Logistic regression analysis demonstrated statistically significant protective effect of ACD over PACG (P < 0.01). The odds ratio corresponding to an increase of 0.01 mm in ACD and LT were computed as 0.83 and 1.11 respectively.

Anterior Chamber↗
PubMed1979

VP-16, cyclophosphamide, adriamycin and cis-platinum (V:CAP-I) in patients with metastatic adenocarcinoma of the lung.

In an attempt to improve upon the 42% regression rate of the CAP-I regimen in patients with advanced adenocarcinoma of the lung, VP-16 was added to that regimen. VP-16, as a single agent, had a response rate of 12.5% (3/24) In a similar group of patients. The new regimen, V:CAP-I, had a tumor regression rate of 35% (7/20) and an estimated median survival of 171 days. Hence, we were unable to conclude that the addition of VP-16 to the CAP-I regimen statistically improved the regression rate of the CAP-I regimen.

Adenocarcinoma↗
PubMed2001

Long-term trends in childhood diabetes mortality: 1968-1998.

OBJECTIVE: In the context of recent improvements in type 1 diabetes therapy, to describe longitudinal trends in mortality attributable to childhood diabetes and to investigate socioeconomic and health services correlates of mortality. RESEARCH DESIGN AND METHODS: We extracted mortality data for 1968-1998 from National Center for Health Statistics files and covariates from the Bureau of Health Professions Area Resource File. Analytical techniques included linear and Poisson regression and standard descriptive statistics. RESULTS: Childhood (defined as 0-19 years of age) age-adjusted mortality from diabetes declined from 9.5 (1968) to 3.0 (1984) deaths per 10 million but remained relatively constant subsequently. All-cause childhood mortality, however, continued to decline. Older children experienced higher mortality rates, as did those living in counties with higher levels of unemployment. CONCLUSIONS: Despite recent improvements in therapy, diabetes-related mortality among children has not declined for 14 years. This finding may be partially attributable to sociodemographic factors influencing access to care, but the remaining mortality may defy available treatment methods. Reducing childhood diabetes mortality rates below the current apparent plateau may require new prevention and/or treatment strategies.

Adolescent↗
PubMed2005

Propensity score methods gave similar results to traditional regression modeling in observational studies: a systematic review.

OBJECTIVE: To determine whether adjusting for confounder bias in observational studies using propensity scores gives different results than using traditional regression modeling. METHODS: Medline and Embase were used to identify studies that described at least one association between an exposure and an outcome using both traditional regression and propensity score methods to control for confounding. From 43 studies, 78 exposure-outcome associations were found. Measures of the quality of propensity score implementation were determined. The statistical significance of each association using both analytical methods was compared. The odds or hazard ratios derived using both methods were compared quantitatively. RESULTS: Statistical significance differed between regression and propensity score methods for only 8 of the associations (10%), kappa = 0.79 (95% CI = 0.65-0.92). In all cases, the regression method gave a statistically significant association not observed with the propensity score method. The odds or hazard ratio derived using propensity scores was, on average, 6.4% closer to unity than that derived using traditional regression. CONCLUSIONS: Observational studies had similar results whether using traditional regression or propensity scores to adjust for confounding. Propensity scores gave slightly weaker associations; however, many of the reviewed studies did not implement propensity scores well.

Bias↗
PubMed1998

Artificial neural networks and logistic regression as tools for prediction of survival in patients with Stages I and II non-small cell lung cancer.

The prognosis of patients with Stage I and II non-small cell lung cancer (NSCLC) can be estimated but cannot be definitively ascertained by use of current clinicopathologic criteria and tumor marker studies. The potential value of probabilistic neural networks (NNs) with genetic algorithms and multivariate logistic regression to predict the survival of NSCLC patients has not been previously evaluated. Multiple prognostic factors (age, sex, cell type, stage, tumor grade, smoking history, and immunoreactivity to c-erbB-3, bcl-2, Glut1, Glut3, retinoblastoma gene and p53 were correlated with 5-year survival in 63 patients with Stage I or II NSCLC, treated solely by surgical excision at Baylor Medical College, Houston, Texas. Several probabilistic NNs with genetic algorithm models were developed using the prognostic features as input neurons and survival at 5 years (free of disease/dead of disease) as output neurons. The probabilistic NN yielded excellent classification rates for dependent variable survival. The best model was trained with 52 cases and classified all 11 "unknown" test cases correctly. Several statistically significant logistic regression models were fitted using 50 cases to build the models and 13 cases as "hold-out" test cases. These multivariate statistical models provide various cutoff values that predict/classify the probability of survival at 5 years. In conclusion, probabilistic NNs and logistic regression models can be useful in estimating the prognosis of patients with Stage I and II NSCLC using multiple clinicopathologic and molecular variables. These multivariate predictive models need to be validated with much larger groups of patients to assess their potential clinical value.

Adult↗
PubMed2001

Cyclin D1 polymorphism and risk for squamous cell carcinoma of the head and neck: a case-control study.

A G-->A polymorphism (G870A) in exon 4 of the cyclin D1 (CCND1) gene creates an alternative splice site in its mRNA, encoding a protein with an altered C-terminal domain. It has been suggested that DNA damage in cells with the A allele bypasses the G(1)/S checkpoint of the cell cycle more easily than damage in cells without the A allele. Because CCND1 plays a critical role in cell cycle control and reduced DNA repair capacity is associated with an increased risk for squamous cell carcinoma of the head and neck (SCCHN), we hypothesize that this CCND1 polymorphism modulates individual susceptibility to SCCHN. To test this hypothesis we evaluated the frequency of the polymorphism in a hospital-based case-control study of 233 newly diagnosed SCCHN patients and 248 non-cancer controls. The cases and controls were frequency matched by age (+/-5 years), sex and tobacco use. All subjects were non-Hispanic whites. We found that the A allele frequency was slightly higher in the cases (0.485) than in the controls (0.425), but the difference was borderline statistically significant (P = 0.064). The frequencies of the CCND1 AA, GA and GG genotypes were 23.6, 49.8 and 26.6%, respectively, in cases and 16.5, 52.5 and 31.5%, respectively, in controls. Multivariate logistic regression analysis adjusting for age (in years), sex, smoking and alcohol use was performed to calculate odds ratios (OR) and 95% confidence intervals (CI). Compared with the wild-type CCND1 GG, the CCND1 A G genotype was associated with a non-significantly increased risk (adjusted OR 1.15, 95% CI 0.75-1.76), but the CCND1 AA genotype was associated with a significantly increased risk (adjusted OR 1.77, 95% CI 1.04-3.02) for SCCHN. Results from a trend test using a logistic regression model were statistically significant (P = 0.044). Among the cases the mean age of onset was 59.0, 56.8 and 55.5 years for the GG, GA and AA genotypes, respectively. In the stratification analysis the CCND1 AA variant genotype was associated with a >3-fold increased risk in individuals who were </=50 years old (OR 3.18, 95% CI 1.19-8.46), females (3.57, 1.26-10.0), non-smokers (3.71, 1.37-10.1) and non-alcohol users (4.76, 1.61-14.0). These results suggest that the CCND1 polymorphism is associated with early onset of SCCHN and contributes to susceptibility to SCCHN in this population.

Carcinoma, Squamous Cell↗
PubMed2006

Head rotation during vertical impact predicted using initial head angle and anthropometry.

INTRODUCTION: Dynamic inertial loading to the head and neck complex, similar to what is experienced during the rocket boost phase of pilot ejection, results in diverse kinematic responses when observed in live human volunteers. The purpose of this study was to develop a predictive model of head rotation during the compressive phase of pilot ejection. METHODS: Post hoc analyses were conducted on data from two previous experimental studies. This analysis included observing 138 tests involving 27 human volunteers (both men and women) where the various kinematic responses were categorized into five modes based on the location of primary rotation and the direction of head rotation. Several statistical methods (logistic regression, linear regression, and Fisher's Exact test) were employed to evaluate the influence of independent variables, body anthropometry, and initial head angle on dependent variables head acceleration and direction of head rotation. RESULTS: Statistical results of this data indicated that initial head position and body anthropometry are significant factors with respect to head response during the compressive phase of an ejection. Two statistical tools were developed; one to assist in predicting the location of primary head rotation, and the other to predict the direction of head rotation. DISCUSSION: By using the two statistical tools together, a simplified method was developed for predicting the modes of head response to vertical impact based on initial position and anthropometry. The results of this study show the importance of initial position prior to an ejection and may assist in identifying individuals at greater risk of injury during an ejection.

Anthropometry↗
PubMed2004

Human papilloma virus testing knowledge and attitudes among women attending colposcopy clinic with ASCUS/LGSIL pap smears.

OBJECTIVE: To study women's knowledge regarding the role of human papilloma virus (HPV) in cervical intraepithelial neoplasia and their attitudes toward the integration of HPV testing as part of routine follow-up of atypical squamous cell of uncertain significance/low-grade squamous intraepithelial lesion (ASCUS/LGSIL) abnormalities. METHODS: Over a 12-month period, all women attending the University of Ottawa colposcopy clinic for evaluation and follow-up of ASCUS/LGSIL Pap smears were recruited. Demographic data included age, nature of the Pap smear abnormality, gravidity, parity, occupation and education level, smoking history, previous history of abnormal smears, colposcopic examination and treatment, and current method of contraception. The women were asked to rate their level of concern over their Pap smear abnormality, from 0 (not concerned) to 10 (very concerned). Women's knowledge regarding the role of HPV in cervical intraepithelial neoplasia and the rationale behind the use of HPV testing was assessed by the clinic nurse as being minimal, moderate, or good, as defined by pre-specified criteria. Upon explanation by the nurses of the results of the recent ALTS (ASCUS/LGSIL Triage Study) trial, the women were asked to state whether they preferred to continue with regular colposcopic surveillance every 6 months, or to use the results of the HPV test, if negative, to reduce the number of colposcopy examinations to one annually. Descriptive statistics and logistic regression analysis were used to identify significant demographic factors associated with the women's preference for incorporation of HPV testing in their follow-up. All P values less than.10 were considered to be statistically significant, due to the exploratory nature of the study. RESULTS: Of the 100 women who participated in the study, 42% presented with ASCUS. The mean age (+/- SD) of the women was 33.63 +/- 11.25 years (range, 18-75 years); 66% were office workers with at least a community college degree, 86% reported previous abnormal Pap smears, and 67% had previously been seen for colposcopy. Fifty-eight percent of the women rated their concern level as being 6 or more, while 15% ranked their concerns as maximal at 10. In terms of knowledge about HPV, 75% of the women had no or very minimal knowledge of the role of HPV in cervical intraepithelial neoplasia. With regard to HPV testing, 84% of the study group had either never heard of the test or had only a minimal knowledge of HPV testing. After being informed of the ALTS results, 64% of the women chose to use HPV testing to help in triaging the needs for frequent colposcopy. Logistic regression modelling showed that a college level education (odds ratio [OR], 2.27; 95% confidence interval [CI], 0.95-;5.45; P =.06) and history of previous treatment for abnormal Pap smears (OR, 3.31; CI, 0.88- 12.46; P =.07) were closely associated with the adoption of HPV testing in clinical management. CONCLUSION: There exists a significant lack of knowledge about HPV and its role in the pathogenesis of cervical intraepithelial neoplasia. Women who have received previous treatments for cervical intraepithelial neoplasia and those with college-level educations were more likely to adopt this new technology as part of their care.

Adolescent↗
PubMed2000

[Analyses of factors prolonging the length of hospital stay in elderly patients beginning hemodialysis].

The aim of the present study is to clarify the factors causing prolongation of the length of hospital stay in elderly patients beginning hemodialysis. Patients aged over 60 years who had newly started hemodialysis (98 cases) were studied. These were 59 men and 39 women. The age was 73 +/- 7 years (mean +/- standard deviation). In each patient, the cause of renal failure (non-diabetes/diabetes), body mass index, comorbid conditions (cerebrovascular disease, ischemic heart disease, etc.), ambulation, cognitive function, urgency of the initiation of dialysis, occurrence of access failure, marital status, younger cohabitants, and the length of stay after initiation of dialysis were surveyed. The median and the mean of the length of stay were 37 and 49 days. Because of this disparity, a normal distribution of the length of stay could not be obtained. However, the distribution was transformed to close to normal by logarithmic conversion of the number of days. We used the log-converted value as the length of stay for statistical analyses. We investigated the influence of the differences of each factor on the length of stay. The subjects were divided into two groups for each factor. The mean and standard deviation of the length of stay was calculated respectively. Comparisons were carried out by unpaired t-test. Multiple regression analysis was also performed using background factors as explanatory variables, and the length of stay as a dependent variable. The factors presented by the nominal scale were converted to dummy variables. Eight variables in the unpaired t-test and seven variables in multiple regression analysis were statistically significant. All but one variable were common to both analyses. The gender was statistically significant only in the unpaired t-test. It could be explained by close correlation of gender with marital status. Access failure and urgent initiation of dialysis were dominant factors for the prolongation of the length of stay. Ischemic heart disease, diabetes, inability to walk, impaired cognitive function, and absence of a partner also prolonged the length of stay.

Aged↗
PubMed1995

An unusual outbreak of food poisoning.

On August 25 1990, over 400 people who attended a Thailand handicappeds' sport day at a provincial physical education college developed gastrointestinal symptoms after having dinner. An epidemiological team want to determine causes(s) and recommend how to prevent and control a food poisoning outbreak. The investigation included interviewing all 1,210 persons who attended the sport's day. In addition, an environmental survey, laboratory analysis of food samples, and rectal, ear, throat and nasal swabs from foodhandlers were also performed. A case was defined as a person who ate any items of dinner food and experienced vomiting, nausea, abdominal pain, and diarrhea. There were 485 cases out of 1,094 persons, an attack rate of 43%. Interviews were completed for 470 out of 485 cases. The three most common symptoms were nausea (93%), vomiting (88%), and abdominal pain (81.5). The mean incubation period was 3.20 hours. Three out of four items of food had a significant association with illness. Among these 3 items, eclairs had to the highest crude relative risk, 7.0 (95% CI = 4.8, 10.2). For statistical analysis, logistic regression by unconditional method was used, and found that only eclairs which were prepared during the night before the dinner and kept at room temperature for at least 12 hours before serving, remained statistically significant in the model (RR = 11.96; 95% CI = 9-22). Laboratory examination of foods and foodhandlers indicated heavy growth of Staphylococcus aureus producing toxins A and C and Bacillus cereus in eclairs. Culture of nasal swabs from healthy foodhandlers identified B. cereus and S. aureus of different phage types from those in eclairs.(ABSTRACT TRUNCATED AT 250 WORDS)

Adolescent↗