A Two-stage Community-based Screening for Diabetic Neuropathy

• Main Authors: Wei-Chih Hsu, 許維志
• Other Authors: Hou-Chang Chiu
• Format: Others
• Language: en_US
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/47919932229112886331

碩士 === 國立臺灣大學 === 預防醫學研究所 === 90 === I. Abstract Background The introduction of neuropathy in type 2 diabetes mellitus has long been neglected over the past decades partly due to unclear etiology or disease natural history and partly due to the lacking of appropriate screening tool and community-based study for early detection of diabetic neuropathy. Objectives Our goals of this study were to use a two-stage community-based design： (1) To estimate prevalence and adjusted prevalence of diabetic neuropathy in Taiwan (2) To estimate risk factors associated with diabetic neuropathy Materials and Methods Patients with diabetes mellitus for estimating the prevalence rate and identifying significant risk factors associated with diabetic neuropathy in this thesis were derived from a population- based multiple screening project in Keelung denoted as Keelung Integrated Community-based Screening (KCIS)。A total of 3,968 cases with level of blood sugar greater than 126 mg/dl were defined as type 2 diabetes mellitus。In order to test the validity of Neurological Symptom Score (NSS)，a validated sample consisting of 121 cases with type 2 diabetic mellitus received both NSS and nerve conduction velocity (NCV) test。Another 587 type 2 diabetes mellitus cases were selected for conducting main study。Subjects who screened positive were electrophysiological examined by technicians for nerve conduction test。 Potential risk factors were assessed，including fasting plasma glucose， HbA1c，body mass index (BMI)，retinopathy，age，sex，diabetic duration， total cholesterol，triglyceride。Age-adjusted Z score was used for the parameters in electrophysiological studies，including nerve action potentials and conduction velocity。Bayesian Approach for adjusted prevalence rate of diabetic neuropathy was conducted。The E-M algorithm approach was developed to the following tackle the missing on negative results of NSS。 Results The overall prevalence rate was 27.79%。The sensitivity， specificity，positive predictive value and negative predictive value were 73.02%，33.33%，58.97% and 51.52% respectively。The estimates of sensitivity，specificity and positive predictive value were 75%，33.33% and 42.86% for old Type 2 DM cases，and 72.22%, 29.69% and 22.41% for newly diagnosed Type 2 DM cases，respectively。The overall prevalence rates were 40.00% and 21.95% for old cases and newly diagnosed cases。Given Bayesian，the posterior marginal estimates were 68.31% (53.45%-68.52%) for sensitivity and 34.67% for specificity (28.95%-39.84%)。The adjusted prevalence rate was estimated as 31.01% (95% CI: 23.39%-39.97%)。Significant factors accounting for the disparity included age (T=3.17, P=0.0017)，sex (X2=7.8695)，duration (T=4.69)，retinopathy (X2=12.6822)，fasting plasma glucose (T=3.45) and HbA1C (T=3.27) in the overall group that included validated sample and the subjects in the main study。 Conclusion: 31% prevalence rate of diabetic neuropathy was estimated using a two-stage design for screening type 2 diabetes mellitus from a community-based multiple screening project。Significant factors for early detection of diabetic neuropathy included age，gender，duration of diabetes，HbA1c，fasting plasma glucose，and retinopathy。