To evaluate the validity of ICD-9-CM and ICD-10-CM coding algorithms in the scoring of APACHE II, APACHE III, and SAPS II in patients admitted to an intensive care unit (ICU).

1. To evaluate the accuracy of automated electronic APACHE II, APACHE III, and SAPS II scoring derived from variations of ICD-9-CM and ICD-10-CM code sets against chart reviews in ICU patients.
2. To assess which ICD-9-CM and ICD-10-CM code set for APACHE II, APACHE III, and SAPS II scoring would be most reliable and practical for use in retrospective cohort studies.

This will be a retrospective observational study. Patients will be identified by randomly selecting from stratified groups to ensure an equal number of patients with ICD-9-CM and ICD-10-CM codes.

• Patients 18 years or older
• Patients admitted to an ICU between 7/1/2014 and 6/30/2016

• Patients with incomplete data
• Pregnant patients
• Prisoners

• Patient Demographics
  • Age
  • Gender
  • Weight
  • Race
  • Type of admission
    • Scheduled surgical
    • Medical
    • Unscheduled surgical
  • ICU length of stay
  • Discharge disposition
• Laboratory (minimum and maximum value during the first 24 hours of the first ICU admission)
  • Basic metabolic panel
    • Sodium
    • Potassium
    • Bicarbonate
    • Creatinine
    • BUN
    • Glucose
  • Liver function tests
    • Albumin
    • Bilirubin
  • Arterial Blood Gas
    • Arterial pH
    • PaCO2
    • PaO2
  • Venous Blood Gas
    • VO2 Saturation
    • VpO2
    • VpCO2
    • Venous pH
    • Bicarbonate
  • Complete Blood Count
    • WBC
    • Hematocrit
  • Urine output (total recorded value for in the first 24 hours of ICU admission)
• Vital signs (minimum and maximum value during the first 24 hours of ICU admission)
• Oxygenation (minimum and maximum value during the first 24 hours of ICU admission)
  • Respiratory rate
  • Mechanical ventilation or CPAP
  • FiO2
  • PaO2
  • PaCO2
• Cardiac arrest (within 24 hours prior to ICU admission)
• Glasgow coma score (minimum during the first 24 hours of ICU admission)
  • Glasgow coma score eye
  • Glasgow coma score verbal
  • Glasgow coma score motor
• Diagnosis and procedural coding (all diagnosis and procedural codes for the encounter)
  • ICD-9-CM or ICD-10-CM codes
  • DRG code
• Comorbidities
  • Liver
    • Cirrhosis
    • Upper GI bleeding
    • Hepatic Failure
    • Encephalopathy
    • Coma
  • Cardiovascular
    • Heart failure
  • Respiratory
    • Chronic restrictive, obstructive, or vascular disease
    • Chronic hypoxia
    • Hypercapnia
    • Secondary polycythemia
    • Pulmonary hypertension
    • Respiratory dependency
  • Renal
    • Acute renal failure
    • Receiving chronic dialysis
  • Immunocompromised
    • Metastatic cancer
    • Immunosuppression
    • Chemotherapy
    • Radiation
    • Long-term or high-dose steroids
    • Leukemia
    • Multiple myeloma
    • Lymphoma
    • AIDS

1. The difference in score between electronic and manual scoring for each of APACHE II, APACHE III, and SAPS II.

1. The difference in predicted mortality between electronic and manual scoring for each risk score.
2. The difference in number of comorbidities identified by electronic and manual scoring for each risk score.
3. The number of comorbidities identified by electronic or by manual scoring but not by both.

1. Analysis for ASHP abstract submission
2. Analysis for ASHP poster
3. Supplemental Material for ASHP Poster