Background
: When research based evidence is not sufficient, clinical practice guidelines can be based on opinions. In such situations, formal consensus development methods, often based on the modified nominal group techniques are widely used. It can be used to evaluate consistency, generalizability, applicability of recommendation when evidence comes from other countries.

Methods : To develop evidence based guidelines for osterarthritis pharmacotherapy, a consensus expert panel consisting of internists, family physicians, methodologists, and orthopedic surgeons were convened. After an extensive structured literature searching and evaluation, evidence statements for key question were developed. Rating methods for consistency, generalizability, applicability of statement were adopted from those jointly developed by Rand and the University of California, Los Angeles.

Results : We developed 27 evidence statements in 17 question domains. Among 72 rating items, 62 items reached agreement. Among 15 recommendations, 10 recommendation grading were A, 2 were B, and 3 were C.

Conclusion : When research based evidence is not sufficient, clinical practice guidelines can be based on formal consensus of experts, especially modified nominal group techniques. It can be used to evaluate consistency, generalizability, applicability of recommendation when evidence comes from other countries.

Background
: Papanicolaou smear adequacy is directly related to endocervical edll discovery, and therefore can improve the value of the Papanicolaou smear as a cancer screening test. One of the pruposes of this study is to assess capture rate of endocervical element of the current Pap smear method in Korea which is performed by using only dry cotton swab. The other purpose is to compare effectiveness of cytobrush cell collector when performing Pap smear.

Methods : This study was performed for 333 patients visiting Family medicine outpatient department at Konkuk University Medical College Hospital to take Pap smear test from March 3, 1994 to July 12, 1994. The both dry cotton swab and cytobrush cell collector were used to collect specimen for each of total 333 patients.

Results : The endocervical cell capture rate was 37.2% for dry cotton swab, 73.0% for cytobrush. The difference of endocervical cell capture rates among two methods was significant statistically. A significant higher number of atypical epithelial changes was found in smears with endocervical cells than in smears without endocervical epithelial changes was found in smears with endocervical cells than in smears without endocervical columnar cells. Thus, the chance of missing an abnormal epithelial change is increased in smears without endocervical columnar cells irrespective of instruments. The use of the cytobrush incresased the number of smears that contained endocervical cells for reproductive age and postmenopausal women.

Conclusion : The use of cytobrush is considered as a more effective method for collection of endocervical cells compared with dry cotton swab, and the chance of detecting an abnormal epithelial change is increased in smears with endocervical cells.