TY - JOUR
T1 - Validity of Diagnostic Algorithms for Inflammatory Bowel Disease in Japanese Hospital Claims Data
AU - Takahashi, Sayumi
AU - Obara, Taku
AU - Kakuta, Yoichi
AU - Shimoyama, Yusuke
AU - Naito, Takeo
AU - Moroi, Rintaro
AU - Kuroha, Masatake
AU - Shiga, Hisashi
AU - Kinouchi, Yoshitaka
AU - Masamune, Atsushi
N1 - Funding Information:
We thank the past and present members of the IBD group at Tohoku University Hospital for their fruitful discussions and scientific contributions; Tatsuro Ishikawa of Tohoku University Hospital for his assistance in preparing the administrative data; and Jodi Smith, from Edanz (https://jp.edanz.com/ac, accessed on 23 May 2022) for editing the draft of this manuscript.
Publisher Copyright:
© 2022 by the authors. Licensee MDPI, Basel, Switzerland.
PY - 2022/7/1
Y1 - 2022/7/1
N2 - Inflammatory bowel disease (IBD) diagnoses are increasing in Japan. Some patients have symptoms that are difficult to control, and further research on IBD is needed. Claims databases, which have a large sample size, can be useful for IBD research. However, it is unclear whether the International Classification of Diseases, Tenth Revision (ICD-10) codes alone can correctly identify IBD. We aimed to develop algorithms to identify IBD in claims databases. We used claims data from the Department of Gastroenterology, Tohoku University Hospital from 1 January 2016 to 31 December 2020. We developed 11 algorithms by combining the ICD-10 code, prescription drug, and workup information. We had access to the database which contains all the information for Crohn’s disease and ulcerative colitis patients who visited our department, and we used it as the gold standard. We calculated sensitivity, specificity, positive predictive value (PPV), and negative predictive value for each algorithm. We enrolled 19,384 patients, and among them, 1012 IBD patients were identified in the gold standard database. Among 11 algorithms, algorithm 4 (ICD-10 code and ≥1 prescription drugs) showed a strong performance (PPV, 94.8%; sensitivity, 75.6%). The combination of an ICD-10 code and prescription drugs may be useful for identifying IBD among claims data.
AB - Inflammatory bowel disease (IBD) diagnoses are increasing in Japan. Some patients have symptoms that are difficult to control, and further research on IBD is needed. Claims databases, which have a large sample size, can be useful for IBD research. However, it is unclear whether the International Classification of Diseases, Tenth Revision (ICD-10) codes alone can correctly identify IBD. We aimed to develop algorithms to identify IBD in claims databases. We used claims data from the Department of Gastroenterology, Tohoku University Hospital from 1 January 2016 to 31 December 2020. We developed 11 algorithms by combining the ICD-10 code, prescription drug, and workup information. We had access to the database which contains all the information for Crohn’s disease and ulcerative colitis patients who visited our department, and we used it as the gold standard. We calculated sensitivity, specificity, positive predictive value (PPV), and negative predictive value for each algorithm. We enrolled 19,384 patients, and among them, 1012 IBD patients were identified in the gold standard database. Among 11 algorithms, algorithm 4 (ICD-10 code and ≥1 prescription drugs) showed a strong performance (PPV, 94.8%; sensitivity, 75.6%). The combination of an ICD-10 code and prescription drugs may be useful for identifying IBD among claims data.
KW - claims database
KW - diagnostic algorithm
KW - inflammatory bowel disease
KW - validity
UR - http://www.scopus.com/inward/record.url?scp=85132864179&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85132864179&partnerID=8YFLogxK
U2 - 10.3390/ijerph19137933
DO - 10.3390/ijerph19137933
M3 - Article
C2 - 35805591
AN - SCOPUS:85132864179
SN - 1661-7827
VL - 19
JO - International Journal of Environmental Research and Public Health
JF - International Journal of Environmental Research and Public Health
IS - 13
M1 - 7933
ER -