BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211207T055411Z
LOCATION:223
DTSTART;TZID=America/Chicago:20211114T170000
DTEND;TZID=America/Chicago:20211114T171500
UID:submissions.supercomputing.org_SC21_sess434_ws_cafcw114@linklings.com
SUMMARY:Electronic Health Records (EHR) Significantly Under-Capture Patien
 t Co-Morbidity
DESCRIPTION:Workshop\n\nElectronic Health Records (EHR) Significantly Unde
 r-Capture Patient Co-Morbidity\n\nDilling, Howard, Ansley\n\nBackground:\n
 The concept of a digital twin in healthcare is predicated upon mimicking, 
 as closely as possible, the clinical state of the patient. However, EHR im
 plementations might suffer in data quality, as they are incumbent upon acc
 urate/complete data entry (typically) into discrete data fields. Unfortuna
 tely, the well-documented problem of healthcare provider “click fatigue” m
 ight preclude full capture of data. \n\nOne of the most important predicto
 rs of cancer outcome is performance status, which also conceptually incorp
 orates the patient’s comorbid illnesses. Does the EHR potentially underest
 imate patients’ disease burden? In this analysis, we compared ICD-10 captu
 re of patient co-morbidities from the EHR’s Diagnoses with those from the 
 billing system (coded by billers from each clinic note).\n\nWe hypothesize
 d that discrete comorbidity data from the billing system would be statisti
 cally larger in number and scope than those captured within the EHR.\n\nMa
 terials/Methods:\nWe utilized a cohort of lung cancer patients treated wit
 h radiotherapy who were covered under a retrospective research protocol. C
 omorbidity information was separately pulled from the EHR (PowerChart, Cer
 ner Medical Systems, Kansas City, MO) and the billing (B) system (Soarian,
  Cerner Medical Systems, Kansas City, MO). Cohorts were standardized, util
 izing only patients with data available from both systems. ICD-9 codes, SN
 OMED codes and invalid ICD-10 codes were removed. Because chronic conditio
 ns were recapitulated at each patient visit in B, all duplicates were remo
 ved. By doing so, we generated a “maximal” comorbidity list for each patie
 nt from each system for purposes of comparison. \n\nCharlson Comorbidity I
 ndex (CCI) and Elixhauser (EL) scores (validated instruments of comorbidit
 y) were generated for each patient from these ICD-10 codes (CCI-EHR, CCI-B
 , EL-EHR, and EL-B, respectively). The CCI captures morbidity across  17 s
 pecific health sub-domains and the EL captures 31. Mean EHR and B scores a
 cross each sub-domain were compared for both CCI and EL using Welch’s t-te
 st. Lastly, the EHR and B data sets were merged to generate CCI-Combined a
 nd EL-Combined scores, which were statistically compared across sub-domain
 s to the CCI-B and EL-B results in turn. All analyses were performed in R 
 version 4.03 using the ‘comorbidity’ package to generate the CCI/EL scores
 .\n\nResults:\nAfter cleaning, EHR and B contained 1929 and 4179 distinct 
 ICD-10 codes, respectively, across 2059 patients. 2582 of these codes were
  exclusive to B. Mean scores across the CCI sub-domains were typically gre
 ater by an order of magnitude or more in B. Interestingly “cancer” was doc
 umented more frequently in the EHR (p=0.19) and “AIDS” was numerically mor
 e common in B, but not significantly so (p=0.058). EL scores were vastly s
 ignificantly higher in B compared with EHR, except for blood loss anemia (
 p=0.083). Comparing CCI-Combined and EL-Combined to the CCI-B and EL-B dat
 a sets statistically improved four sub-domain scores -- cerebrovascular di
 sease (p< 2.2e-16) and renal disease (p<2.2e-16) in CCI and hypothyroidism
  (p<2.2e-16) and solid tumor diagnosis (p=9.6e-9) in EL.\n\nConclusion:\nT
 he EHR significantly underestimates patient co-morbidity. If billing data 
 are available, they should be incorporated into patient co-morbidity calcu
 lations.\n\nTag: Applications, Computational Science, Education and Traini
 ng and Outreach, HPC Community Collaboration, HPC Training and Education, 
 Machine Learning and Artificial Intelligence, Performance, Workforce\n\nRe
 gistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR