Updated CharlsonIndex.sql per #160 #161
Conversation
Codecov Report
@@ Coverage Diff @@
## develop #161 +/- ##
============================================
+ Coverage 23.28% 93.47% +70.18%
============================================
Files 13 15 +2
Lines 1095 1165 +70
============================================
+ Hits 255 1089 +834
+ Misses 840 76 -764
Continue to review full report at Codecov.
|
|
Maybe easier to maintain if a reference table is created (like a csv with fields comorbidityId, conceptId). Store the csv file in the package path, read to memory as data frame, bulk load to db, perform inner join? Reason - may make the maintaining easier. The advantage of the old approach of using ancestor_concept_id + descendants was it was more resistant to vocabulary changes. This change, that involves specifying each of the descendantConceptIds is more susceptible to vocabulary changes. |
| WHERE descendant_concept_id IN (4160276, 443392, 4131422, 4131428, 4131304, 4130839, | ||
| 4131938, 4133007, 4130842, 443252, 439392, 193144, 4312802, | ||
| 78097, 4246450, 4246451, 378087, 4312290, 140960, 4247962, | ||
| 192568, 4312023, 200959, 439751, 196053, 200348, 46273652, | ||
| 4281027, 198700, 44806773, 254591, 318096, 442182, 320342, | ||
| 434298, 434875, 373425, 199752, 72266, 4147162, 253717, | ||
| 4315806, 196925, 46270513, 4281030, 136354, 198371, 4314071, | ||
| 4158910, 78987, 432851) |
There was a problem hiding this comment.
@sfortin1 this list includes 439392, Primary malignant neoplasm. is that supposed to be an excluded concept?
New CCI coding algorithm per #160.