Skip to main content

Table 4 Association of pain category (combining self-report with EHR data) with 30-day rehospitalization: TRACE-CORE, 2011–2013

From: Non-specific pain and 30-day readmission in acute coronary syndromes: findings from the TRACE-CORE prospective cohort

Pain condition Incidence of 30-day rehospitalization Model 1a: Unadjusted Model 2b: Adjusted for baseline characteristics Model 3c: Model 2 further adjusted for self-reported chest pain
n/N (%) OR (95% CI) P value OR (95% CI) P value OR (95% CI) P value
No pain 21/285 (7.4) Reference   Reference   Reference  
Mild to moderate pain, documented in EHR 9/101 (8.9) 1.23 (0.54–2.78) 0.62 1.23 (0.52–2.90) 0.63 0.99 (0.42–2.40) 1.00
Mild to moderate pain, not documented in EHR 46/347 (13.3) 1.92 (1.12–3.30) 0.02* 2.03 (1.14–3.62) 0.02* 1.56 (0.84–2.90) 0.16
Severe pain 11/54 (20.4) 3.22 (1.45–7.14) 0.004* 3.11 (1.30–7.42) 0.01* 2.57 (1.05–6.29) 0.04*
  1. Patient-self-reported pain status was collected for the first month after discharge from index hospitalization (using SF-36 survey distributed at one month post-discharge). Electronic health record (EHR) documentation of pain was extracted by natural language processing from patient’s clinical notes created within 30 days post discharge and before readmission to the hospital (if any)
  2. TRACE-CORE indicates Transitions, Risks, and Actions in Coronary Events Center for Outcomes Research and Education
  3. aModel 1: logistic regression, unadjusted for covariates
  4. bModel 2: multivariable logistic regression, adjusted for age, sex, smoking status, acute coronary syndrome type, comorbidity, patient-self-reported generic pain, and healthcare seeking behaviors reported at index hospitalization
  5. cModel 3: Model 2 further adjusted for patient-report of chest pain, tightness, or angina on the Seattle Angina Questionnaire
  6. *Indicates statistically significant (P < 0.05)