Font Size: a A A

Analysis Of Features On Non-skeletal Phase And Establishment Of An Early Discrimination Model Of Skeletal Fluorosis Exposed To Water-Drinking

Posted on:2012-08-04Degree:MasterType:Thesis
Country:ChinaCandidate:S HanFull Text:PDF
GTID:2154330332996601Subject:Epidemiology and Health Statistics
Abstract/Summary:PDF Full Text Request
Objective(1) To assess the diagnostic value by the clinical symptoms for skeletal fluorosis exposed to water-drinking,with X-ray diagnosis for gold standard.(2) To establish and evaluate an early discrimination model of this disease.Method(1) Select two Fluorosis areas in a county and make a definite diagnosis by X-ray of aged 30 and above; make the clinical diagnosis according to symptoms were detected, Determine the content of Ca, Cu, Fe, Mg, P, Zn in the peripheral blood, urine and ALP,γ-GT, LDH, ALT, CHE; compare the differences of the indicators between different groups.(2) A case control study by distinguishing patients and non-patients and analyzing the indexes of skeletal fluorosis.(3) the discriminant analysis method from the metrological diagnostics; summarize the classification regularity by mathematical theory and statistical analysis method, according to the data and information of some sample in the same category, establish the discriminant formula and regularity of the criterion named the discriminant model. cross-validation method for the evaluation.(4) Statistical method: Analyze of variance,χ2 test and discriminant analysis by SPSS 16.0.Results(1) The most common manifestations are pain and stiff of bone and joint, the accurately rates of clinical and X-ray were 69.3% and 63.2%,the difference between the results of the two methods was not statistically significant (P>0.05); the coincidence rate was 69.3%, the cross was appeared in the sub-degree; the sensitivity was 80.5 %, the specificity was 50.0%, the false negative rate was 19.5%, the false positive rate was 50.0%.(2) 467 cases were detected,194 cases were patient and 273 were non-patient. The results show: ward,age, gender, income,education,received fluorine years,The staple food is the locally grown ,type of drinking water; the difference between the two groups are statistically significant. Except numbness and convulsions, the two populations bone clinical symptoms (including pain, stiff, and physical activity obstacles and the difference of body deformation) are statistically significant. There were differences between the patients and non-patient groups of chemical elements in serum Ca,Mg,P,Zn content (P<0.05). Patients group was higher than non-patient group;as well as,contents of serum Ca, P,Zn in different levels of skeletal fluorosis patients were significant different (P<0.05). In this way, serum Ca,P content increased. With the severity increasing of skeletal fluorosis;serum Ca decreased svevrer skeletal fluorosis patients. ALP andγ-GT were significantly different. Not only in different populations from fluorosis areas,but also in different levels of skeletal fluorosis patients (P<0.05). Urine Ca contents which from the fluorosis endemic patients were significantly higher than normal population (P<0.05). In addition, the significant differences among the varying X-ray degrees of populations about urinary Calcium content did show (P<0.05). Besides,urinary fluoride content were not significantly different(P>0.05).(3) The risk analysis of endemic skeletal fluorosis under different exposure showed that: ward, age,and access fluorine years,drinking water of village wells may be risk factors;women,higher education and high income may be potential protective factors of the disease; pain, stiffness, impairment and other clinical manifestations of limb deformity possibility of suffering from the disease who were not present the symptoms 1.957, 1.920, 2.305, 3.215 times.(4) Diagnosis of skeletal fluorosis by received fluorine years,the area under ROC curve is 0.834, P = 0.000; so it has high diagnostic value. Different yesrs as a diagnostic sector have different sensitivity and specificity. Proposed to take received fluorine years is 25 years as a screening of endemic skeletal fluorosis best cutoff value (sensitivity 92%, specificity 25%); when the diagnostic cutoff value set at 36 years, is the best cutoff level for sensitivity and specificity (sensitivity 80.9%, specificity 60.2%).(5) Fisher stepwise discriminant analysis method tested six indicators including the following equation: received fluorine years (X1),ward (X2),ALP (X3),stiffness (X4),drinking water type (X5) and culture degree (X6);the early diagnosis model of endemic skeletal fluorosis by drinking water is as follows: Non-patient group: Y = -27.991 + 6.197 X1 + 0.120 X2 - 0.400 X3 + 0.215 X4 + 8.934 X5 +0.963 X6; Patient group: Y = -34.205 + 7.797 X1 + 0.121 X2 - 0.538 X3 + 0.335 X4 + 8.386 X5 + 1.057 X6;Assessment results of cross validation:The discriminate accuracy of the discriminate model for the non-patient group was 81.7%, the patients rate was 82.6%.And the other results are: the sensitivity was 82.6% ,the specificity was 81.7%,the missed diagnosis rate was 17.4%, the misdiagnosis rate was 18.3%,the correct index of 0.643, positive likelihood ratio of 4.5, negative likelihood ratio was 0.2, compliance rate was 82% and Kappa was 0.634; we can see that the discriminate model is effective for the early screening and diagnosis of skeletal fluorosis.(6) Bayes discriminant analysis can establish the model as follows: Non-patient group: Y= -28.284 + 6.318 X1 + 0.121 X2 - 0.393 X3 + 0.218 X4 + 8.806 X5 + 0.834 X6; Mild patient group: Y= -33.000 + 7.454 X1 + 0.121 X2 - 0.277 X3 + 0.316 X4 + 8.321 X5 + 1.044 X6; Moderate and severe patient group: Y= -35.696 + 8.371 X1 + 0.124 X2 - 0.653 X3 + 0.352 X4 + 8.092 X5 + 0.834 X6;Cross validation evaluation results show that the correct identification rate of discriminate model for non-patient group was 76.2%, assessment accuracy for the mild patient group was 38.5%, assessment rate for moderate and severe patient group was 86.8%; which shows the model of non- patients and patients with moderate to severe discrimination is better.Conclusion(1) The relevance ratio of clinical and X-ray diagnosis is consistent in Endemic fluorosis, the specificity of the diagnosis by the clinical symptom for the disease is low, the false positive rate are high, maybe due to the non-objectivity of the clinical symptom, and it should to be controlled. The results of the clinical and X-ray diagnosis for this disease are consistent. Although the way cannot be separately used,it can provide important clues.(2) Pain and stiffness are the only early clinical symptoms of skeletal fluorosis, with ward, age, gender, income, education, received fluorine years, the staple food is the locally grown and the type of drinking water can be used as meaningful indicators of the early diagnosis of endemic skeletal fluorosis. ALP might be significant to endemic fluorosis inducing liver injury in the pathogenesis of early diagnosis;(3) Skeletal fluorosis patients received fluoride yesrs was significantly higher than non-patients, Determination of the received fluoride yesrs can help the early diagnosis of the disease;(4) For skeletal fluorosis, there is no an ideal and non-invasive method for early diagnosis. Analysis by general, clinical manifestations, blood in urine biochemistry and enzymology of population, and use discriminate analysis to develop the discriminant model can improve the early diagnosis rate of the disease.
Keywords/Search Tags:Skeletal fluorosis Exposed to Water-Drinking, clinical, X-ray, early diagnosis, discriminant analysis
PDF Full Text Request
Related items