Table 2.
Informative features associated with HNSCC by HPV status as discovered by TEPAPA.
| Log (OR) | P | N | Text feature | Type | EMR Source | Interpretation | Crossref. |
|---|---|---|---|---|---|---|---|
| Informative features associated with HPV-related HNSCC | |||||||
| 3.50 | 3.0 × 10−6 | 25 | “HPV (studies|genotypes|status):? P16 immunohistochemistry:? Positive” | R | Pathology | HPV status (Self-referent) | (S3c.1) |
| 3.89 | 6.2 × 10−6 | 20 | “HPV (positive|genotypes: Positive|associated squamous cell carcinoma|related).” | R | Pathology | HPV status (Self-referent) | (S3c.2) |
| 3.29 | 2.0 × 10−5 | 23 | “No FDG avid? pulmonary (nodules|nodule) or pleural” | R | PET | (Lack of) metastasis to the lung | (S4c.1) |
| 3.14 | 5.6 × 10−5 | 21 | “HPV related” | N | Pathology | HPV status (Self-referent) | (S3b.6) |
| 2.06 | 0.00094 | 24 | “irradiation (and|with) (or without|concurrent) chemotherapy” | R | MDT | Management | (S2c.7) |
| 2.76 | 0.0093 | 9 | “oropharyngectomy:” | N | Pathology | Management, site of primary tumor | (S3a.22) |
| 3.23 | 0.0011 | 13 | “SCC of the right (tonsil|base of tongue|glossotonsillar sulcus) -” | R | MDT | Site of primary tumor | (S2d.4) |
| 2.68 | 0.0015 | 16 | “SCC of the (right|left)? base of tongue” | R | MDT | Site of primary tumor | (S2d.5) |
| 3.02 | 0.0023 | 11 | “M0” | N | MDT | Stage | (S2a.3) |
| 2.89 | 0.0047 | 10 | “non-keratinising” | N | Pathology | Pathology feature | (S3a.16) |
| 2.77 | 0.0092 | 9 | “p16? positive,? HPV? positive” | R | MDT | HPV status (Self-referent) | (S2c.17) |
| Informative features associated with HPV-unrelated HNSCC | |||||||
| −3.54 | 0.00035 | 8 | “for decalcification” | N | Pathology | Pathology feature | (S3e.2) |
| −2.91 | 0.00089 | 10 | “a (locally|locoregionally)? (p16 negative|advanced) SCC” | R | MDT | HPV status and pathology feature | (S2h.3) |
| −3.17 | 0.0031 | 6 | “SCC of the supraglottic? (lower lip|larynx).” | R | MDT | Site of primary tumor | (S2g.7) |
| −2.96 | 0.0086 | 5 | “likely to? require adjuvant radiation therapy” | R | MDT | Management | (S2g.10) |
| −3.35 | 0.0011 | 7 | supportive care | N | MDT | Management | (S2f.3) |
| −2.59 | 0.0058 | 8 | “differentiated, keratinising squamous cell carcinoma” | N | Pathology | Pathology feature | (S3e.23) |
| −2.59 | 0.0058 | 8 | “well differentiated” | N | Pathology | Pathology feature | (S3e.26) |
Note: The type field indicates the type of text features (N: n-gram fragments or R: regular expression). N indicates number of documents containing the text features. Abbreviations: Log (OR): Log odds ratio. MDT: Multidisciplinary team meeting.