The Validity and Reliability of Korean Version of Lubben Social Network Scale / 가정의학회지

BACKGROUND: Over the past decade, considerable evidence has been accumulated showing that social networks influence health. However few valid and reliable instruments exist for assessing social networks among elderly population. Therefore, we translated the Lubben social network scales (LSNS) and tested the validity and reliability of Korean version. METHODS: Ten items, self-administered questionnaires were translated by multidisciplinary committee members. Also they were reverse translated by bilingual and modified from pretest procedure. Reliability was tested by 4 weeks testretest reliability and internal consistency (Cronbach's alpha). Validity was tested by factor analysis and construct validity comparison with the Korean version of Duke-UNC and Korean Health Related Quality of Life Scale (KQOLS) were tested for validity testing. RESULTS: Cronbach's alpha was 0.75. Four weeks test-retest Pearson coefficient was 0.78 (P < 0.0001). Correlation coefficient with the Korean version of Duke-UNC was 0.58 (P < 0.0001). Correlation coefficient with social function domain of KQOLS was 0.40 (P < 0.0001). Principal components analysis identified 3 factors. Three contents areas are family and friends networks, interdependent social support and living arrangements. CONCLUSION: Korean version of LSNS is both valid and reliable instrument, but fine refinements will be needed to adapt for use in practical settings.

The Quality of Reporting of Randomized Controlled Trials in Korean Medical Journals Indexed in KoreaMed: Survey of Items of the Revised CONSORT Statement

BACKGROUND: The revised Consolidated Standards for Reporting of Trials (CONSORT) were developed to improve the reporting of Randomized Controlled Trials. We studied to survey the extent to which RCTs report items included in the revised CONSORT recommendations. METHODS: A descriptive survey of RCTs enrolled in 2005 at KoreaMed, which is a representative database in Korea was done. The main outcome measures were the proportion of RCTs that reported each of 22 checklist items of CONSORT. RESULTS: We identified 125 RCTs from 26 journals. Random sequence implementation (0%), estimated effect size and its precision (0%), sample size determination (8.9%), method of random sequence generation (7.3%), allocation concealment (3.2%), participant flow (4.8%) and any other analysis (7.3%), generalizability of the trial findings (0.8%) were pooly reported. CONCLUSION: The proportions of following the CONSORT recommendations in Korean medical journals were very low. An effort to improve the reporting of RCTs by application and recommendation of CONSORT statement is required.

Evidence Based Evaluation of Health Information in the Television News

BACKGROUND: A rapid growth of socioeconomic status in Korea has triggered health information resolution of unprecedented magnitude among the general population. Despite its obvious benefits, the increase in the amount of information could also result in many potentially harmful effects on both consumers and professionals who do not use it appropriately. METHODS: This study was performed to evaluate health information in the television news from June 2003 through January 2004. We evaluated the quality and the accuracy of health information provided in the night news. We reviewed the health information for quality using the evidence-based medicine tools which evaluate the accuracy required to understand the text. RESULTS: A total of 85 types of information were identified. Among them, 34 (40.0%) contained inaccurate or misleading statement based on evidence-based medicine. These included confusing surrogate outcome with an end outcome (15.3%), extrapolating nonhuman results to human (8.2%), exaggerating results in conclusion (12.9%), incorrect words (7.1%), indefinite study methods (2.4%) In broadcasting stations A, B and C, the rate of error were 55.6%, 38.2% and 42.4%, respectively. The rate of error were higher in information of Korean source compared to those of international sources (65.8% and 27.3%). CONCLUSION: Many inaccurate medical information exist in the television news.

The Usefulness of Waist/Height Ratio as an Obesity Index

BACKGROUND: Waist circumference and BMI have been known as an independent predictor for cardiovascular diseases. But some people with a normal BMI or normal waist circumference may have cardiovascular risk factors. Therefore, the purpose of this study was to investigate the simplicity and the usefulness of waist/height ratio as an index of obesity and to investigate the waist/height ratio as a predictor for metabolic risks in individuals with normal weight and/or normal waist circumference. METHODS: The subjects were 1,157 individuals (635 men, 522 women). The morbidity index for metabolic risk factors was calculated by the sum of the risk factors scores (one point per item, if present). We set the criteria of obesity as BMI>or=25 kg/m2, waist circumference of >or=90 cm in males and >or=80 cm in females, waist/hip ratio of >or=0.9 in males and >or=0.8 in females, and waist/height ratio of >or=50 cm/cm*100. RESULTS: There was a significant correlation among waist/ height ratio, waist/hip ratio, BMI, and waist circumference (P=0.000). The odds ratios for morbidity index (>or=3) in normal BMI, waist circumference, and waist/hip ratio subjects with waist/height over 50 were significantly higher with 4.60, 3.83, and 2.79, respectively, and also higher in normal BMI with normal waist circumference subjects with 3.90 (P=0.000). CONCLUSION: Waist/height ratio may be used a simple and practical index of obesity. Especially, it may be useful in subjects with normal BMI or normal waist circumference when predicting their metabolic risks.