Core Needle Biopsy Rate for New Breast Cancer Diagnosis
Core Needle Biopsy Rate for New Breast Cancer Diagnosis
Simple descriptive statistics such as mean ± standard deviation, median/range, and percentages were reported. Sex and age were identified at the time of the first occurrence of breast cancer and were analyzed per patient. All other analyses were performed per unique cancer identified, with some patients presenting with synchronous lesions, as there was felt to be adequate independence between cancer diagnoses to consider each malignancy diagnosed as a single analysis unit. A binomial proportion test was used for the analysis of leftversus right-sided tumors to test for an unequal proportion. A twosample t test was used for comparison of the mean number of tissue acquisition attempts for the categories of 14- versus 18-gauge needle biopsy and for CNB versus CNB with vacuum assistance. Cases were excluded from this analysis when there was a combination of 14- and 18-gauge needles used or if the number of attempts was not known. When analyzing clinical tumor size estimates, the one-sample t test was used to compare the reported pre- and/or postbiopsy size difference, and the two-sample t test was used to compare the reported pre-/post-CNB size estimates for stereotactic-guided and ultrasoundguided CNB. Cases with more than one biopsy type or with unknown or unreported tumor size were excluded. For the analysis of timeliness, records with an unknown number of days for a given timeliness interval (pathology report to surgery, pathology report to surgery appointment, or surgery appointment to day of surgery) were excluded.
Patients in the study years 2007 and 2008 who had a histologically benign initial needle biopsy, but no immediate surgical biopsy, had follow-up through March, 2011, to judge the success or failure of the CNB (descriptive statistics for follow-up duration are reported as well). Those patients with a histologically malignant needle biopsy all had a positive surgical biopsy. Using this information, standard performance metric point estimates of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were produced along with respective 95% confidence intervals; when the point estimates were 1 the Clopper-Pearson exact intervals were used, otherwise the Wald limits were used. A simple Kappa agreement statistic was calculated as well; corresponding 95% confidence interval was also calculated. Definitions were as follows: sensitivity = TP/(TP + FN), specificity = TN/(TN + FP), PPV = TP/(TP + FP), and NPV = TN/(TN + FN).
Statistical Analysis
Simple descriptive statistics such as mean ± standard deviation, median/range, and percentages were reported. Sex and age were identified at the time of the first occurrence of breast cancer and were analyzed per patient. All other analyses were performed per unique cancer identified, with some patients presenting with synchronous lesions, as there was felt to be adequate independence between cancer diagnoses to consider each malignancy diagnosed as a single analysis unit. A binomial proportion test was used for the analysis of leftversus right-sided tumors to test for an unequal proportion. A twosample t test was used for comparison of the mean number of tissue acquisition attempts for the categories of 14- versus 18-gauge needle biopsy and for CNB versus CNB with vacuum assistance. Cases were excluded from this analysis when there was a combination of 14- and 18-gauge needles used or if the number of attempts was not known. When analyzing clinical tumor size estimates, the one-sample t test was used to compare the reported pre- and/or postbiopsy size difference, and the two-sample t test was used to compare the reported pre-/post-CNB size estimates for stereotactic-guided and ultrasoundguided CNB. Cases with more than one biopsy type or with unknown or unreported tumor size were excluded. For the analysis of timeliness, records with an unknown number of days for a given timeliness interval (pathology report to surgery, pathology report to surgery appointment, or surgery appointment to day of surgery) were excluded.
Patients in the study years 2007 and 2008 who had a histologically benign initial needle biopsy, but no immediate surgical biopsy, had follow-up through March, 2011, to judge the success or failure of the CNB (descriptive statistics for follow-up duration are reported as well). Those patients with a histologically malignant needle biopsy all had a positive surgical biopsy. Using this information, standard performance metric point estimates of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were produced along with respective 95% confidence intervals; when the point estimates were 1 the Clopper-Pearson exact intervals were used, otherwise the Wald limits were used. A simple Kappa agreement statistic was calculated as well; corresponding 95% confidence interval was also calculated. Definitions were as follows: sensitivity = TP/(TP + FN), specificity = TN/(TN + FP), PPV = TP/(TP + FP), and NPV = TN/(TN + FN).
Source...