Ikuko Kato, Geogia Spivak, Jim Collins, Seongho Kim, Jeffrey Duncan, Jun Sun
{"title":"Statewide Linkage Study of <i>Salmonella</i> Infection and Colorectal Cancer Incidence in Michigan, United States.","authors":"Ikuko Kato, Geogia Spivak, Jim Collins, Seongho Kim, Jeffrey Duncan, Jun Sun","doi":"","DOIUrl":null,"url":null,"abstract":"<p><strong>Objective: </strong>Nontyphoidal <i>Salmonella</i> infection is one of the most common foodborne illnesses, and its oncogenic potential has been documented in animal models. The primary goal of this study was to examine whether individuals who were exposed to enteric <i>Salmonella</i> infection are more likely to develop colorectal cancer (CRC) than the general population through the linkage of 2 statewide public health surveillance databases.</p><p><strong>Materials and methods: </strong>We designed a 2-stage probabilistic linkage, starting with 17,587 records of enteric salmonellosis reported to Michigan Department of Health and Human Services between 1992 and 2020. These records did not include unique identifiers (such as Social Security number [SSN]). The initial linkage to LexisNexis address history was conducted to obtain information to calculate each person's time in Michigan as well as SSN for the second linkage. The linkage to the state cancer registry was performed to obtain the observed number of CRC cases, while the expected number of CRC cases was calculated according to corresponding state CRC incidence by age, sex, and calendar year.</p><p><strong>Results: </strong>Ninety-three percent of the initially identified salmonellosis records were sent to LexisNexis linkage, which returned address history, death, and SSN for 97% of the records. Further linkage to the statewide cancer registry identified 98 incident CRC cases. Overall, the observed-to-expected (O/E) ratio was not different from unity (0.833; 95% CI, 0.627-1.003).</p><p><strong>Conclusions: </strong>While the new linkage strategy was found effective and should be applicable to other health conditions, we cannot rule out bias due to incomplete or underreporting of the infection in estimating the risk of CRC.</p>","PeriodicalId":39246,"journal":{"name":"Journal of registry management","volume":"51 2","pages":"62-68"},"PeriodicalIF":0.0000,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11343435/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of registry management","FirstCategoryId":"1085","ListUrlMain":"","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"Medicine","Score":null,"Total":0}
引用次数: 0
Abstract
Objective: Nontyphoidal Salmonella infection is one of the most common foodborne illnesses, and its oncogenic potential has been documented in animal models. The primary goal of this study was to examine whether individuals who were exposed to enteric Salmonella infection are more likely to develop colorectal cancer (CRC) than the general population through the linkage of 2 statewide public health surveillance databases.
Materials and methods: We designed a 2-stage probabilistic linkage, starting with 17,587 records of enteric salmonellosis reported to Michigan Department of Health and Human Services between 1992 and 2020. These records did not include unique identifiers (such as Social Security number [SSN]). The initial linkage to LexisNexis address history was conducted to obtain information to calculate each person's time in Michigan as well as SSN for the second linkage. The linkage to the state cancer registry was performed to obtain the observed number of CRC cases, while the expected number of CRC cases was calculated according to corresponding state CRC incidence by age, sex, and calendar year.
Results: Ninety-three percent of the initially identified salmonellosis records were sent to LexisNexis linkage, which returned address history, death, and SSN for 97% of the records. Further linkage to the statewide cancer registry identified 98 incident CRC cases. Overall, the observed-to-expected (O/E) ratio was not different from unity (0.833; 95% CI, 0.627-1.003).
Conclusions: While the new linkage strategy was found effective and should be applicable to other health conditions, we cannot rule out bias due to incomplete or underreporting of the infection in estimating the risk of CRC.