Quality Assessment of Non-Randomized Studies in the Korean Journal of Family Medicine.
10.4082/kjfm.2009.30.2.129
- Author:
Hyung Joon KIM
1
;
Soo Young KIM
Author Information
1. Department of Family Medicine, Kangdong Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Korea. hallymfm@gmail.com
- Publication Type:Original Article ; Comparative Study
- Keywords:
Korean Journal of Family Medicine;
Non-Randomized trial;
Quality Assessment
- MeSH:
Humans;
Korea;
Research Design
- From:Korean Journal of Family Medicine
2009;30(2):129-137
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
BACKGROUND: Since the Korean Journal of Family Medicine (KJFM) started publishing in 1980, randomized controlled trials (RCTs) and non-randomized controlled studies (NRSs) consistently have increased in quality and quantity. Although there have been several studies on the quality assessment of RCTs in Korea, there has been no study on quality assessment of NRSs. Thus, this study evaluated NRS in the KJFM to assess the quantity and quality. METHODS: Upon extracting NRSs for assessing the intervention effects from all the articles published in the KJFM from 1980 to 2006, assessments were made on methodological index for non-randomized studies (MINORS). Also, the analysis were made upon the proportion of NRSs within original articles according to two categories (comparative study, noncomparative study). The mean scores by research methods and years, and total scores and mean scores of yearly research methods were analyzed. RESULTS: A total of 34 trials on NRSs were selected. In 1980s, 0.65% of the total selected original articles, in 1990s, 1.54%, and in 2000s 5.11% were NRSs. According to the research designs, the mean scores of MINORS were before and after study 8.5, interrupted Time Series 9.7 (out of 16), controlled before and after 13.5, quasi randomized trial 12.6, and non randomized trial 13 (out of 24). Before and after design was the most frequently used (55.9%). CONCLUSION: Although NRSs consistently increased in quantity, the assessed mean scores were low and most articles used before and after design. Thus, there should be studied using appropriate research methods in the future.
