Natural Language Processing (NLP) offers an automated method to extract data from unstructured free text fields for arthroplasty registry participation. Our objective was to investigate how accurately NLP can be used to extract structured clinical data from unstructured clinical notes when compared with manual data extraction. A group of 1,000 randomly selected clinical and hospital notes from eight different surgeons were collected for patients undergoing primary arthroplasty between 2012 and 2018. In all, 19 preoperative, 17 operative, and two postoperative variables of interest were manually extracted from these notes. A NLP algorithm was created to automatically extract these variables from a training sample of these notes, and the algorithm was tested on a random test sample of notes. Performance of the NLP algorithm was measured in Statistical Analysis System (SAS) by calculating the accuracy of the variables collected, the ability of the algorithm to collect the correct information when it was indeed in the note (sensitivity), and the ability of the algorithm to not collect a certain data element when it was not in the note (specificity).Aims
Methods
We report on 397 consecutive revision total hip
replacements in 371 patients with a mean clinical and radiological follow-up
of 12.9 years (10 to 17.7). The mean age at surgery was 69 years
(37 to 93). A total of 28 patients (8%) underwent further revision,
including 16 (4%) femoral components. In all 223 patients (56%,
233 hips) died without further revision and 20 patients (5%, 20
hips) were lost to follow-up. Of the remaining patients, 209 (221
hips) were available for clinical assessment and 194 (205 hips)
for radiological review at mean follow-up of 12.9 years (10 to 17.7). The mean Harris Hip Score improved from 58.7 (11 to 92) points
to 80.7 (21 to 100) (p <
0.001) and the mean Merle d’Aubigné and
Postel hip scores at final follow-up were 4.9 (2 to 6), 4.5 (2 to
6) and 4.3 (2 to 6), respectively for pain, mobility and function.
Radiographs showed no lucencies around 186 (90.7%) femoral stems
with stable bony ingrowth seen in 199 stems (97%). The survival
of the S-ROM femoral stem at 15 years with revision for any reason as
the endpoint was 90.5% (95% confidence interval (CI) 85.7 to 93.8)
and with revision for aseptic loosening as the endpoint 99.3% (95%
CI 97.2 to 99.8). We have shown excellent long-term survivorship and good clinical
outcome of a cementless hydroxyapatite proximally-coated modular
femoral stem in revision hip surgery. Cite this article:
