Reliability of medical records in diagnosing inflammatory breast cancer in Egypt.

BACKGROUND

Inflammatory breast cancer (IBC) is a rare, aggressive breast cancer diagnosed clinically by the presence of diffuse erythema, peau d'orange, and edema that arise quickly in the affected breast. This study evaluated the validity of medical records in Gharbiah, Egypt in identifying clinical signs/symptoms of IBC. For 34 IBC cases enrolled in a case-control study at the Gharbiah Cancer Society and Tanta Cancer Center, Egypt (2009-2010), we compared signs/symptoms of IBC noted in medical records to those recorded on a standardized form at the time of IBC diagnosis by clinicians participating in the case-control study. We calculated the sensitivity and specificity of medical records as compared to the case-control study for recording these signs/symptoms. We also performed McNemar's tests.

RESULTS

In the case-control study, 32 (94.1%) IBC cases presented with peau d'orange, 30 (88.2%) with erythema, and 31 (91.2%) with edema. The sensitivities of the medical records as compared to the case-control study were 0.8, 0.5, and 0.2 for peau d'orange, erythema, and edema, respectively. Corresponding specificities were 1.0, 0.5, and 1.0. p values for McNemar's test were <0.05 for all signs. Medical records had data on the extent and duration of signs for at most 27% of cases for which this information was recorded in the case-control study. Twenty-three of the 34 cases (67.6%) had confirmed diagnosis of IBC in their medical records.

CONCLUSIONS

Medical records lacked information on signs/symptoms of IBC, especially erythema and edema, when compared to the case-control study. Deficient medical records could have implications for diagnosis and treatment of IBC and proper documentation of cases in cancer registries.