Erratum: Enabling qualitative research data sharing using a natural language processing pipeline for deidentification: Moving beyond HIPAA Safe Harbor identifiers (JAMIA Open (2021) 4:3 (ooab069) DOI: 10.1093/jamiaopen/ooab069)

Abstract

In the originally published version of this manuscript, there were inaccuracies in the frequencies of identifiers. In fourth paragraph of RESULTS, the following should read: "The number of identifiers varied in our 2 datasets but ranged between 1-3% of the total word tokens (Table 3). These identifiers are located within large volumes of unstructured qualitative text, making manually locating them challenging."instead of: "The number of identifiers varied in our 2 datasets but ranged between 0.01% and 0.03% of the total word tokens (Table 3). The overall frequency of identifiers is relatively low due to the large volume of unstructured qualitative text within which identifiers are located."And in the second paragraph of DISCUSSION, the following should read: "We found very few HSH and non-HSH identifiers in qualitative text (1-3% of all words)."instead of "We found very few HSH and non-HSH identifiers in qualitative text (0.01-0.03% of all words)."Frequencies were affected in the Identifier Count column of Table 3 also. This should read:.