BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//conferences.acspri.org.au//2022//talk//WRMC7E
BEGIN:VTIMEZONE
TZID:AEST
BEGIN:STANDARD
DTSTART:20000326T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3;UNTIL=20050326T170000Z
TZNAME:AEST
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
END:STANDARD
BEGIN:STANDARD
DTSTART:20060402T040000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060401T170000Z
TZNAME:AEST
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
END:STANDARD
BEGIN:STANDARD
DTSTART:20070325T040000
RRULE:FREQ=YEARLY;BYDAY=4SU;BYMONTH=3;UNTIL=20070324T170000Z
TZNAME:AEST
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
END:STANDARD
BEGIN:STANDARD
DTSTART:20080406T040000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4
TZNAME:AEST
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000827T030000
RRULE:FREQ=YEARLY;BYDAY=4SU;BYMONTH=8;UNTIL=20000826T170000Z
TZNAME:AEDT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20011028T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20071027T170000Z
TZNAME:AEDT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20081005T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=10
TZNAME:AEDT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-2022-WRMC7E@conferences.acspri.org.au
DTSTART;TZID=AEST:20221124T140000
DTEND;TZID=AEST:20221124T141500
DESCRIPTION:Background: Hospital admission records contain a rich resource 
 of data for healthcare research\, providing a direct insight into processe
 s and procedures whilst also being resilient to bias and limitations affli
 cting other sampling methods. In Australian hospitals\, most data records 
 are standardised or otherwise classified using internationally established
  conventions (e.g.\, International Classification of Disease by the World 
 Health Organisation)\, thereby providing a robust data source for research
 . Hospital admission records are not centrally stored\, with emergency and
  inpatient datasets located separately with different structures and frame
 works. Therefore\, before utilising hospital records data to report outcom
 es\, pre-processing steps need to be taken. Here\, we homogenise and link 
 emergency and inpatient admission datasets and apply natural language proc
 essing on the linked datasets to create a predictive model for patient len
 gth of stay and readmission.\n\nMethods: The dataset contains emergency an
 d inpatient hospital admission records from two local health districts (So
 uth-Eastern-Sydney and Illawarra-Shoalhaven Local Health Districts) betwee
 n 2020 and 2021. Both datasets were configured to a patient-admission leve
 l by reshaping the datasets to have the diagnostic records expand across a
  row rather than a column. A custom algorithm was created to link the resh
 aped datasets by using de-identified patient IDs as key and matching overl
 apping admission and departure/discharge date-times. Two outcome variables
  were generated for natural language processing: one indicating if the pat
 ient was readmitted within 28 days\, and another indicating if the patient
  was admitted for more than one day. Diagnostic records from the emergency
  dataset\, inpatient dataset\, as well as age and gender of the patient we
 re used in the models to predict the outcomes based on natural language pr
 ocessing (random forest classification with TF-IDF and word2vec vectors). 
 Stata MP 15.1 was used to pre-process the datasets\, and Python was used t
 o link the datasets at a patient-admission level and run the natural langu
 age processing algorithm. The study was conducted under ethical approval f
 rom South-Eastern Sydney Local Health District Human Research Ethics commi
 ttee (HREC/16/POWH/412) and Macquarie University\, and funded under a Nati
 onal Health and Medical Research Council Partnership Project (1111925).\n\
 nResults: Without the emergency dataset linked\, the TF-IDF model produced
  a predictive model for readmission with 96% precision and 76.2% recall. T
 he linkage increased the precision to 96.4% and the recall to 76.3%. The u
 nlinked word2vec model had a precision of 96.7% and a recall of 74.8%\, wh
 ich increased to 97.1% precision and slightly reduced to 74.6% recall afte
 r linkage. For predicting if the patient would be admitted for more than o
 ne day\, the unlinked TF-IDF model had 86.2% precision and 89.7% recall\, 
 which increased to 87.4% precision and 90.5% recall after linkage. The unl
 inked word2vec model had 81.9% precision and 89% recall\, and the linked m
 odel had 83.4% precision and 88.6% recall.\n\nConclusion: Hospital admissi
 on records provide a rich source of data for secondary data analysis\, wit
 h pre-processing and linking different components of a patient’s stay imp
 roving predictive modelling. Here we show an improvement in predictive mod
 elling by linking inpatient and emergency dataset diagnostic records. Link
 age with pathology tests\, radiology tests\, and medications would further
  improve predictive models and reporting outcomes.\n\nRecording link: http
 s://acspri-org-au.zoom.us/rec/share/sFS4E0Eva6L3CETeEbh9cr8bLW7jHf3BqadVrD
 3ZBgTXkIpMYjtO_h9UWipxxHi-.ZbxNTlh6Oe1hAEZ5?startTime=1669255431000
DTSTAMP:20260615T202021Z
LOCATION:Zoom Breakout Room 3
SUMMARY:Linking Hospital Emergency and Inpatient Admissions for secondary d
 ata analysis: a case study using Natural Language Processing - Gorkem Sezg
 in
URL:https://conferences.acspri.org.au/2022/talk/WRMC7E/
END:VEVENT
END:VCALENDAR