Table 5.

Agreement regarding type and purpose of colorectal endoscopy for Medicare claims, medical records, and self report

Test type by source
Test purpose by source
Data sources compared*SIG in both sourcesSIG in source 1 COL in source 2COL in source 1 SIG in source 2COL in both sourcesAgreement on type of test (95% CI)Screening in both sourcesScreening in source 1 nonscreening in source 2Nonscreening in source 1 screening in source 2Nonscreening in both sourcesAgreement on purpose of test (95% CI)
Claim to medical record (n = 137)42918493% (88-97%)170604852% (43-61%)
31%6%1%62%14%0%48%38%
Self report to medical record (n = 107)†197126882% (75-89%)552410765% (55-74%)
18%7%11%64%57%25%11%7%
Self report to claim (n = 124)†2113157577% (70%-85%)188621729% (20-36%)
17%11%12%60%15%70%1%14%
  • NOTE: Limited to study participants with a single test in the “gold” source and evidence of a test in the second source.

    Abbreviations: SIG, sigmoidoscopy; COL, colonoscopy.

  • * Source 1 refers to the first source listed, and source 2 refers to the second source listed in each comparison. The second source is considered the “gold.”

  • n's do not sum to total because of missing information as to type or purpose of test in medical record or survey.