SEARCH · PubMed Health

Results for “Logistic Models”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 919 records · Page 51Linked to original sources

PubMed1998

Reporting of industrial accidents in The Netherlands.

The true number of industrial accidents followed by absence from work in the Netherlands in 1993 has been estimated by using a capture-recapture approach. A government safety register noted 29,685 cases reported by Dutch companies. The statutory organization that insures employees of the same companies against absence from work for medical reasons (but does not provide special compensation for disability due to accidents at work) noted 19,397 cases for which absence control visits revealed an occupational injury to be the cause of the absence. These two registers showed an overlap of 5,881 matched cases. Assuming them to be independent and capture probability to be homogeneously distributed, the authors obtained a crude estimate of 97,909 accidents. Two methods for correcting this crude estimate are presented. The first method is logistic modeling of the data using age, sex, branch of industry, number of work sites, size of workforce, and seriousness of injury as explanatory variables to allow for heterogeneity in capture probabilities. Applying the model coefficients of these characteristics to the 43,201 distinct cases noted in both registers, we obtained an estimate of 97,524 accidents. Second, knowledge of the reporting and registration procedures utilized by the two sources was used to obtain an estimate of source bias. Errors in recording date of birth and differences between the two registries precluded matching of a number of overlapping cases, accounting for 16.9% overestimation of the true number, corresponding to 83,283 accidents and a government capture rate of 35.6%.

Absenteeism↗

PubMed1993

Clinical utility of the serum pancreolauryl test in diagnosis and staging of chronic pancreatitis.

Indirect pancreatic function tests are frequently used in the clinical routine as complementary tools for the diagnosis of chronic pancreatitis (CP) because of their noninvasiveness and simplicity. We analyzed the clinical efficacy and routine application of a modified serum pancreolauryl test (PLT) in the diagnosis and staging of CP. We studied a total of 90 patients with CP diagnosed by endoscopic retrograde pancreatography and 54 patients with extrapancreatic gastrointestinal disorders as controls. Sensitivity and specificity of the serum PLT in the diagnosis of CP were 82% and 91%, respectively, using a value of 4.5 micrograms/ml as cutoff. In the diagnosis of patients with mild to moderate morphological changes of CP, the sensitivity of the serum PLT (52%) was improved by the concomitant analysis of serum pancreatic amylase in a logistic model (70%). Serum PLT closely correlated with the degree of pancreatic ductal abnormalities (p < 0.001), and showed a sensitivity of 81% and specificity of 89% in the staging of CP (mild-moderate vs. marked CP; cutoff 2.5 micrograms/ml). We conclude that the modified serum PLT is a reliable test which should be considered as a first-line option for the diagnosis and follow-up of patients with CP.

Amylases↗

PubMed2002

Effects of gamma-ray and high energy carbon ion irradiation on swimming velocity of Euglena gracilis.

The effects of gamma-ray and high energy carbon ion irradiation on the swimming velocity of the photosynthetic flagellate Euglena gracilis strain Z were studied, focusing on a dose-effect relationship. Cells were exposed to 60Co gamma-rays at 6 doses of 10, 15, 20, 40, 100 and 200 Gy for water, and also to 290 MeV/amu carbon ions from the Heavy Ion Medical Accelerator in Chiba at 7 doses (5, 10, 15, 20, 50, 100 and 200 Gy for water). The swimming velocity was measured by a biomonitoring system, called ECOTOX. The swimming velocities of Euglena gracilis cells were significantly decreased by >40 Gy gamma-rays and >5 Gy carbon ions, respectively. The 50% effective doses for inhibition, 34 +/- 4 Gy (gamma-rays) and 13 +/- 1 Gy (290 MeV/amu carbon ions), were estimated from the best fit to data of the logistic model. The relative biological effectiveness (2.6 +/- 0.4) was calculated by the ratio of 50% effective doses. The inhibition of the swimming velocity of the cells irradiated with gamma-rays was still present after 3 days, while recovery of the swimming velocity was shown in the cells exposed to 290 MeV/amu carbon ions. It is suggested that ionizing radiation inhibits ATP production and/or increases frictional drag on beating of the flagellum, thus decreasing swimming velocity.

Adenosine Triphosphate↗

PubMed1998

Clinical findings with implications for genetic testing in families with clustering of colorectal cancer.

BACKGROUND: Germ-line mutations in DNA mismatch-repair genes (MSH2, MLH1, PMS1, PMS2, and MSH6) cause susceptibility to hereditary nonpolyposis colorectal cancer. We assessed the prevalence of MSH2 and MLH1 mutations in families suspected of having hereditary nonpolyposis colorectal cancer and evaluated whether clinical findings can predict the outcome of genetic testing. METHODS: We used denaturing gradient gel electrophoresis to identify MSH2 and MLH1 mutations in 184 kindreds with familial clustering of colorectal cancer or other cancers associated with hereditary nonpolyposis colorectal cancer. Information on the site of cancer, the age at diagnosis, and the number of affected family members was obtained from all families. RESULTS: Mutations of MSH2 or MLH1 were found in 47 of the 184 kindreds (26 percent). Clinical factors associated with these mutations were early age at diagnosis of colorectal cancer, the occurrence in the kindred of endometrial cancer or tumors of the small intestine, a higher number of family members with colorectal or endometrial cancer, the presence of multiple colorectal cancers or both colorectal and endometrial cancers in a single family member, and fulfillment of the Amsterdam criteria for the diagnosis of hereditary nonpolyposis colorectal cancer (at least three family members in two or more successive generations must have colorectal cancer, one of whom is a first-degree relative of the other two; cancer must be diagnosed before the age of 50 in at least one family member; and familial adenomatous polyposis must be ruled out). Multivariate analysis showed that a younger age at diagnosis of colorectal cancer, fulfillment of the Amsterdam criteria, and the presence of endometrial cancer in the kindred were independent predictors of germ-line mutations of MSH2 or MLH1. These results were used to devise a logistic model for estimating the likelihood of a mutation in MSH2 and MLH1. CONCLUSIONS: Assessment of clinical findings can improve the rate of detection of mutations of DNA mismatch-repair genes in families suspected of having hereditary nonpolyposis colorectal cancer.

Adaptor Proteins, Signal Transducing↗

PubMed1992

Self-report versus medical record functional status.

The importance of assessing functional status in the hospitalized patient is gaining recognition. However, the availability and accuracy of medical record functional status data are uncertain. We collected data on 2,504 patients greater than 65 years of age discharged alive. A personal interview conducted 2 days before discharge recorded the patient's self-reported ability to perform 5 activities of daily living scales. Medical record abstraction was used independently to determine ability to perform the same activities of daily living scales. Patients who required any human assistance to perform a function were considered dependent. Patients were also contacted after discharge to determine the site of posthospital care (28% discharged to a nursing home). The amount of missing medical record functional status data varied by function from 20% for bathing to 50% for dressing. Ten percent of patients had no medical record functional status documentation concerning any of the five functions. The prevalence of self-reported dependence at discharge varied by function from 24% for feeding to 93% for bathing. The total number of dependencies differed between the two methods (medical records, 2.3 +/- 1.9; self-report data, 3.2 +/- 1.5). There was exact agreement between the two methods on the total number of dependencies in 28% of cases and differences of greater than or equal to 3 in 20%. In a stepwise logistic model predicting discharge to a nursing home and adjusting for other relevant variables, the number of dependencies as determined by self-report and medical record data each remained significant (Odds Ratios = 1.6). Self-report and medical record functional status data differ substantially, and the medical record data remain independently associated with nursing home placement. Several possible explanations for this finding are explored.

Activities of Daily Living↗

PubMed1998

Risk factors for contracting plague in Nkayi district, Zimbabwe.

OBJECTIVE: To identify risk factors for contracting plague among the three to 16 year old age group. DESIGN: A matched case control study, carried out in two parts. SETTING: Nkayi District--a plague endemic area in Zimbabwe. SUBJECTS: A total of 86 cases were randomly selected from a line listing of all reported plague cases in the district. Two neighbourhood controls were selected from households at least 500 m on either side of the case household. MAIN OUTCOME MEASURES: Maximum likelihood estimate of the odds ratio (OR) for the 17 risk factors studied (grouped as personal, environmental, domestic animals and rat reports). RESULTS: Risk factors which were significantly associated with contracting plague were having a sick cat in the household (OR 3.4, 95% CI 1.1 to 13.0), herding cattle (OR 1.7, 95% CI 1.0 to 3.0), and age 10 years or older (OR 2.2, 95% CI 1.2 to 4.1). In a logistic model, only having a sick cat was found to be independently associated with being a case (OR 3.9, 95% CI 1.3 to 12.2). CONCLUSION: The presence of a sick domestic cat was identified as a risk factor for acquiring bubonic plague during this outbreak. Communities should be informed that sick cats may act as a warning of potential plague, and that if cats become sick they should be removed from the household.

Adolescent↗

PubMed2003

[Neighborhood effects influencing reproductive health of Chilean women].

BACKGROUND: Fecundity rates have decreased considerably in Latin America, due to a higher contraceptive use and better family planning programs. AIM: To determine whether neighborhood level socioeconomic variables have an independent effect on reported modern contraceptive use, over and above the effect of individual level measures of socioeconomic status and reproductive health behavior. SUBJECTS AND METHODS: Multilevel logistic models determined the effects of individual and neighborhood dimensions (socioeconomic status, urbanization, quality of public health facilities) on contraceptive use among 509 women aged 15 to 49 years living in 85 neighborhoods within the Region of Bio Bio, Chile. RESULTS: After adjustment for women's individual socioeconomic characteristics and other risk factors, less favorable neighborhood socioeconomic characteristics were significantly associated with lower rates of modern contraceptive use. CONCLUSIONS: Changes in the quality of facilities for family planning at the neighborhood level may enhance modern contraceptive use.

Adolescent↗

PubMed2004

Plasma urocortin in human systolic heart failure.

Urocortin (UCN), a member of the corticotrophin-releasing factor family, is expressed in heart, brain and gut. UCN has potent cardiostimulatory, cardioprotective, vasodilator and diuretic/natriuretic effects, and cardiac UCN expression is increased in heart failure (HF). In the present study, we investigated plasma levels of UCN in 119 patients with HF and 212 age- and gender-matched controls to clarify its relationship with gender and disease severity. UCN was elevated in HF [normal males, 19.5 (3.9-68.8) pmol/l and HF males, 50.2 (6.9-108.2) pmol/l, P < 0.0005; normal females, 14.2 (3.9-53.5) pmol/l and HF females, 21.8 (3.9-112.5) pmol/l, P < 0.001; values are medians (range)]. The relative increase was greater in males than females ( P < 0.03). UCN fell with increasing age, especially in HF patients ( r(s) = -0.56, P < 0.0005) and with increasing New York Heart Association (NYHA) class ( r(s) = -0.55, P < 0.0005). The fall in UCN levels with increasing NYHA class was reinforced by a significant correlation between UCN and ejection fraction ( r(s) = 0.45, P < 0.0005) in HF patients. Although receiver operating characteristic (ROC) curves for diagnosis of all HF cases yielded an area under the curve (AUC) of 0.76, ROC AUCs for patients with early HF (NYHA class I and II) were better (0.91). ROC AUCs for logistic models incorporating N-terminal probrain natriuretic peptide (N-BNP) and UCN were better than either peptide alone. In conclusion, plasma UCN is elevated in HF, especially in its early stages. Its decline with increasing HF severity may expedite disease progression due to diminished cardioprotective/anti-inflammatory effects. UCN measurement may also complement N-BNP in the diagnosis of early HF.

Adult↗

PubMed1999

Using pooled exposure assessment to improve efficiency in case-control studies.

Assays can be so expensive that interesting hypotheses become impractical to study epidemiologically. One need not, however, perform an assay for everyone providing a biological specimen. We propose pooling equal-volume aliquots from randomly grouped sets of cases and randomly grouped sets of controls, and then assaying the smaller number of pooled samples. If an effect modifier is of concern, the pooling can be done within strata defined by that variable. For covariates assessed on individuals (e.g., questionnaire data), set-based counterparts are calculated by adding the values for the individuals in each set. The pooling set then becomes the unit of statistical analysis. We show that, with appropriate specification of a set-based logistic model, standard software yields a valid estimated exposure odds ratio, provided the multiplicative formulation is correct. Pooling minimizes the depletion of irreplaceable biological specimens and can enable additional exposures to be studied economically. Statistical power suffers very little compared with the usual, individual-based analysis. In settings where high assay costs constrain the number of people an investigator can afford to study, specimen pooling can make it possible to study more people and hence improve the study's statistical power with no increase in cost.

Animals↗

PubMed2001

A nonlinear effect of hyperglycemia and current cigarette smoking are major determinants of the onset of microalbuminuria in type 1 diabetes.

Cigarette smoking and poor glycemic control are risk factors for diabetic nephropathy in type 1 diabetes. However, the specifics of the relation of these risk factors to the onset of this complication have not been elucidated. To investigate these issues, we followed for 4 years 943 Joslin Clinic patients aged 15-44 years with type 1 diabetes who had normoalbuminuria during the 2-year baseline period. Microalbuminuria developed in 109 of the 943 individuals, giving an incidence rate of 3.3/100 person-years. The risk of onset of microalbuminuria was predicted somewhat more precisely by the measurements during the 1st and 2nd years preceding onset than by all the measurements during the longer (4-year) interval, suggesting attenuation of the impact of past hyperglycemia over time. Point estimates of the incidence rate (per 100 person-years) according to quartiles of HbA(1c) during the 1st and 2nd years preceding the outcome were 1.3 in the 1st, 1.5 in the 2nd, 3.1 in the 3rd, and 6.9 in the 4th (P = 1.3 x 10(-9)). Point estimates of the incidence rate (per 100 person-years) according to smoking status were 7.9 for current smokers, 1.8 for past smokers, and 2.2 for those who had never smoked (P = 2.0 x 10(-7)). In a multiple logistic model, the independent effects of HbA(1c) level and cigarette smoking remained highly significant, but their magnitudes were reduced. Using the same covariates in a generalized additive model, we examined the shape of the relationship between HbA(1c) and onset of microalbuminuria and found significant nonlinearity in the logarithm of odds scale (P = 0.04). The slope was steeper with HbA(1c) >8% than <8%. Furthermore, the change in the slope was magnified among current smokers. In conclusion, patients with type 1 diabetes who smoke and have an HbA(1c) >8% have the highest risk of onset of microalbuminuria.

Adolescent↗

PubMed1996

[Prognostic factors of Pneumocystis carinii pneumonia in AIDS].

OBJECTIVES: Identify prognosis factors in Pneumocystis carinii pneumonia at diagnosis and construct a model to predict mortality according to these prognosis factors. METHODS: Seventy-seven consecutive cases of proven AIDS-related Pneumocystis carinii pneumonia (67 men, 10 women, mean age 37.2 years) were reviewed to determine the most accurate initial prognostic factors and estimate an individual prediction of death. A stepwise logistic regression analysis was performed. Three kinds of data were entered into the logistic model: historical data, clinical and laboratory data obtained within the first 24 hours of diagnosis, and specific data related to chest X-ray and bronchoalveolar lavage results. RESULTS: The sum of arterial partial pressure of oxygen and carbon dioxide (PaO2 + PaCO2) and serum albumin level best predicted a fatal outcome in multivariate analysis. CONCLUSION: The logistic equation provided by the model might be used to accurately and quickly identify the patients with severe Pneumocystis carinii pneumonia who might benefit from supportive intensive care.

AIDS-Related Opportunistic Infections↗

PubMed2005

Evaluating change in physical activity with the building of a multi-use trail.

BACKGROUND: Cross-sectional studies suggest a positive association between the presence of trails and physical activity participation. Prospective evaluations of the impact of building a multi-use trail, in terms of change in physical activity levels among nearby residents, are needed. DESIGN: The study was designed as a quasi-experimental noncontrol pre-post design. SETTING/PARTICIPANTS: Participants included 366 adults aged > or =18 years living within 2 miles of the evaluated trail. INTERVENTION: A railway of >23 miles was under development for conversion to a multi-use trail in central North Carolina. A segment of the trail was evaluated by randomly selecting and telephone interviewing adults living within 2 miles of the planned trail before trail construction began and approximately 2 months after completion of construction. MAIN OUTCOME MEASURES: Outcomes were time spent in leisure activity, leisure activity near home, walking, bicycling, moderate activity, vigorous activity, and transportation activity. RESULTS: At follow-up, of the 366 adults living within 2 miles of the trail, 11.0% had not heard of the trail, and 23.1% had heard of the trail and had used it at least once. In multivariable logistic models, leisure activity, leisure activity near home, moderate activity, vigorous activity, and walking for transportation did not significantly change for those who used the trail compared to those not using the trail. CONCLUSIONS: This prospective study of the building of a multi-use trail did not demonstrate an increase in physical activity among adults living near the trail. Other prospective studies are encouraged, to take advantage of rigorously evaluating different types of trails that are to be constructed in rural and urban settings.

Adolescent↗

PubMed2004

The social context of childhood injury in Canada: integration of the NLSCY findings.

OBJECTIVE: To integrate findings from cross-sectional and longitudinal analyses of the relationships between childhood injury, child behavior, parenting, family functioning and neighborhood characteristics. METHODS: Logistic modeling of cross-sectional (n = 12,666) and longitudinal (n = 9796) data from the National Longitudinal Survey of Children and Youth. RESULTS: Consistent correlates of childhood injury across designs included child's age, gender, difficult temperament, aggressive behavior, positive parenting, neighbors' cohesion, neighborhood problems, and socioeconomic disadvantage. CONCLUSION: Contextual influences on childhood injury vary by child's age, temperament and behavior. In early childhood, neighborhood processes of cohesion show protective effects. For older children, neighborhood disadvantage dominates the risk of injuries.

Adolescent↗

PubMed2005

Selected characteristics of a new polyvinyl siloxane impression material--a randomized clinical trial.

OBJECTIVES: This study evaluated the ability of a new polyvinyl siloxane impression material (Affinis, Coltène/Whaledent, material A) to obtain final impressions free of bubbles and voids for indirect fixed cuspal-coverage restorations. The results were compared to a control polyvinyl siloxane impression material (material B). Both materials were handled by inexperienced clinicians (undergraduate dental students) in student clinics. METHOD AND MATERIALS: One-hundred and thirty patients who were treated in the Louisiana State University School of Dentistry Junior Student Clinic for indirect fixed cuspal-coverage restorations and who met the inclusion criteria were randomly assigned to either one of two treatment groups, group A (n = 65) or group B (n = 65). Two calibrated examiners evaluated the first impression of prepared posterior teeth at a magnification of 10x for acceptability (no voids or bubbles). Position of tooth, type of preparation, preparation finish line (Class I-V), and gingival bleeding scores were recorded. All statistical tests were performed with the level of significance set at .05. RESULTS: The Fisher-Freeman-Halton test did not reveal significant associations between material and gingival bleeding score (P = .492). Significant differences in the location of the preparation finish line between materials were observed (P = .0096); material A was more frequently used in cases where the preparation finish line was located at least 2 mm subgingivally. Logistic regression was used to assess the effect of the material on the success of the impression (acceptable/ unacceptable). Material was highly significant in the logistic model (P < .001) with an odds in favor of an acceptable impression being eight times higher with material A than with material B (odds ratio = 8.00; 95% confidence index for odds ratio: 2.832, 22.601). The 60/65 (92.3%) impressions made with material A and 39/65 (60%) impressions made with material B were rated "acceptable." CONCLUSION: The new polyvinyl siloxane impression material provided a significantly higher proportion of impressions free of bubbles and voids than the control polyvinyl siloxane material.

Chi-Square Distribution↗

PubMed1993

Discrete subaortic stenosis. Operative age and gradient as predictors of late aortic valve incompetence.

Between January 1969 and May 1990, 100 patients were operated on for discrete subaortic stenosis. Three patients died in the perioperative period. Patients with intrinsic lesions, prosthetic replacement, or extensive operative remodeling of the aortic valve were excluded from the analysis. The 67 remaining patients had a median follow-up of 62 months. Preoperatively, 8 patients had aortic valve competence, 51 had mild incompetence, and 8 patients moderate aortic valve incompetence. At follow-up mild incompetence persisted in 27 and moderate incompetence in 6 patients. In 1 patient it worsened from no incompetence to mild and in another patient from mild to moderate. The probability of aortic incompetence at follow-up was significantly and simultaneously related (multivariate ordinal logistic model) to (1) older age at operation (logarithm of months, p = 0.007), (2) higher preoperative gradient (third power of milligrams of mercury, p = 0.0004), (3) preoperative cardiomegaly (p = 0.04), and (4) surgical myectomy (p = 0.002). There was an interaction between age and gradient (p = 0.03). Two nomograms are proposed as a generalizable aid to decision making. The data support the policy of early repair of subaortic stenosis.

Adolescent↗

PubMed2001

Indomethacin tocolysis and intraventricular hemorrhage.

OBJECTIVE: To determine the association between indomethacin tocolysis and neonatal intraventricular hemorrhage. METHODS: Fifty-six preterm neonates with intraventricular hemorrhage were matched by gestational age with neonates (n = 224) without this morbidity. Maternal and neonatal charts were reviewed to ascertain the type of tocolytic exposure experienced by the neonate. Other maternal and neonatal demographic and outcome data were also abstracted. Results were analyzed using the Student t test, chi(2) analysis, and multivariable logistic regression. The number of studied subjects provided 80% power to determine if antenatal exposure to indomethacin was twice as likely among infants with intraventricular hemorrhage. RESULTS: Univariate analysis revealed that there were no significant differences between the study and control groups with respect to maternal age, parity, or betamethasone exposure. Infants with intraventricular hemorrhage were significantly more likely to be born at an earlier gestational age, a lower birth weight, after maternal chorioamnionitis, after vaginal delivery, and after exposure to either indomethacin alone or a combination of indomethacin and magnesium. Additionally, their neonatal course was significantly more likely to be complicated by sepsis and respiratory distress syndrome. In a multivariable logistic model, only gestational age, chorioamnionitis, vaginal delivery, and respiratory distress syndrome continued to be significantly associated with intraventricular hemorrhage. Indomethacin exposure, either as single-agent (adjusted odds ratio 1.3, 95% confidence interval 0.5, 3.3) or combination tocolytic therapy (adjusted odds ratio 2.0, 95% confidence interval 0.8, 4.8), was not significantly associated with intraventricular hemorrhage. CONCLUSION: Indomethacin tocolysis is not associated with an increased risk of intraventricular hemorrhage.

Adult↗

PubMed2005

Probability of porcine reproductive and respiratory syndrome (PRRS) virus infection as a function of exposure route and dose.

At the most elemental level, the design of effective strategies to control and/or eliminate porcine reproductive and respiratory syndrome (PRRS) virus depend on an accurate and comprehensive understanding of virus transmission. As a general rule, transmission is highly dependent on the route of exposure and the dose of virus. The objective of this study was to derive PRRS virus isolate VR-2332 dose-response curves for oral and intranasal routes of exposure, i.e., determine the probability that a specific virus dose would result in infection. Individually housed pigs approximately 21 days of age were exposed to specific doses of PRRS virus isolate VR-2332 by either oral or intranasal routes. Positive controls were intramuscularly inoculated with 10(2.2) 50% tissue culture infective dose (TCID50) of PRRS virus and negative controls were orally administered 100ml of diluent with no virus. Pigs were monitored for evidence of infection for 21 days following exposure, i.e., serum samples were collected on days 0, 7, 14, 21, and tested for virus and PRRS virus-specific antibodies. Dose-response curves and 95% confidence intervals for oral and intranasal routes of exposure were derived using logistic models (logit and probit). The infectious dose50 (ID50) for oral exposure was estimated to be 10(5.3) TCID50 (95% CI, 10(4.6) and 10(5.9)); the ID50 for intranasal exposure was estimated to be 10(4.0) TCID50 (95% CI, 10(3.0) and 10(5.0)). Given these estimates, it is worth noting that intramuscular exposure of animals to 10(2.2) TCID50 (positive controls) resulted in infection in all animals. Thus pigs were the most susceptible to infection via parenteral exposure.

Administration, Intranasal↗

PubMed1996

Preoperative change of thyroid stimulating hormone receptor antibody level: possible marker for predicting recurrent hyperthyroidism in patients with Graves' disease after subtotal thyroidectomy.

To make the surgical treatment for Graves' disease more ideal, it is important to elucidate factors related to postoperative thyroid dysfunction in addition to thyroid remnant. Because TSH receptor antibody (TRAb) is thought to be one of the essential causes of Graves' disease, we investigated whether preoperative changes in serum TRAb levels are related to postoperative recurrent hyperthyroidism. Between 1987 and 1992 a total of 1520 patients with Graves' disease were treated by subtotal thyroidectomy. Of these patients 335 visited Ito Hospital with no history of drug treatment of their disease and were treated surgically after several courses of antithyroid drug (ATD) therapy. There were 68 males and 267 females with a mean age of 25.8 years. The mean follow-up period was 48 months (range 12-84 months). Factors analyzed by univariate and multivariate analysis were as follows: age, sex, duration of ATD treatment, weight of resected thyroid, weight of thyroid remnant, preoperative titer of MCHA, TRAb at the time of initial examination (TRAb1), TRAb at the time of surgery (TRAb2), and DeltaTRAb, the difference between TRAb1 and TRAb2 (DeltaTRAb = TRAb1 - TRAb2). The chi-square test was used for univariate analysis and a logistic model for multivariate analysis. Of this group, 119 patients were euthyroid (35.5%), 50 were hyperthyroid (14.9%), and 166 were hypothyroid (49.3%). Significant factors related to recurrent hyperthyroidism were weight of thyroid remnant and DeltaTRAb in both univariate and multivariate analyses. DeltaTRAb is a possible new marker for predicting postoperative recurrent hyperthyroidism. If the preoperative TRAb level is not improved by ATDs in patients with Graves' disease, the thyroid remnant should be made smaller.

Adolescent↗