Year: 2018
Type: erratum
Abstract: In the process of working on a subsequent manuscript, an error in the code written to define the MarketScan cohort was identified.Correcting this error has resulted in a lower total number of patients included in the MarketScan cohort, now n ¼ 40 174 compared to 44 344 in the original publication.The reduction in cohort size is primarily due to a lower number of patients treated with mastectomy alone, now n ¼ 6357, compared to 11 432 in the original publication (see Table 1, revised).To reflect these changes, the analyses in the manuscript for the MarketScan cohort have been revised.Figure 1A remains qualitatively similar to the original figure included in the published manuscript.The key finding presented in Table 2, specifically an excess risk of complications with mastectomy and reconstruction compared to lumpectomy and whole breast irradiation, remains essentially unchanged (relative risk [RR] ¼ 1.79, 95% confidence interval [CI] ¼ 1.75 to 1.83; P <.001).The only meaningful change to Table 2 is that treatment with mastectomy alone is no longer associated with a lower risk of any local treatment complication (RR ¼ 1.03, 95% CI ¼ 1.00 to 1.07; P ¼ .09).Table 3 has been updated and key findings remain stable, for example the adjusted mean complication cost is $9708 (95% CI ¼ 8401 to 11 017) with mastectomy and reconstruction and $1336 (95% CI ¼ 1193 to 1479) with lumpectomy and whole breast irradiation.Figure 2A has been updated as well, with key findings qualitatively similar.
DOI: 10.1093/jnci/djy135
Language: English
Subfield: Cancer Research
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W4248190873