Prediction of daily pain and opioid consumption in hospitalized patients with cancer: The importance of psychological symptoms, previous pain, and opioid use.

Abstract

195 Background: Pain is among the most common reasons that cancer patients present to the emergency department (ED) and are ultimately hospitalized, yet little is known about the patient-level predictors and outcomes of pain-related ED visits and hospitalizations. This study sought to evaluate the predictive association of clinical pain and psychological variables on ED admission with average daily pain and opioid consumption in subsequently hospitalized patients with cancer. Methods: A prospective cohort study of patients with active cancer presenting to the ED with pain severity ≥4/10 on presentation completed baseline surveys that assessed demographic, socioeconomic, cancer diagnosis and treatment, medication use, and psychosocial profiles using validated questionnaires. Daily pain scores and opioids administrated were abstracted from the hospital record. Univariable and multivariable general estimating equation (GEE) analyses examined associations of baseline socioeconomic and clinical variables with the primary outcomes, average daily pain and total daily opioid administration. Results: Patients (N = 113) had various types of cancer, with 91% having a diagnosis of > 1 year, 80% with metastatic solid tumors, 73% reported pain as the primary reason for the ED visit, 43% were taking opioids, and 27% had chronic pain predating their cancer. Average daily pain scores ranged between (3.4 to 5.0 out of 10) and daily MMEs ranged between 0 to 1136 MMEs, yet the majority received lower than 60 MMEs per day. In multivariable models, higher pain catastrophizing (B = 0.1, p= 0.001), more recent surgery (B = -0.2 p= 0.016), outpatient opioid use (B = 1.4, p<0.001), and history of chronic pain (B = 0.8, p= 0.022) before cancer diagnosis were independent predictors of greater average daily pain during hospitalization. Higher pain catastrophizing (B = 1.6, p= 0.049), higher anxiety (B = 3.7, p= 0.030), lower depression (B = -4.9, p= 0.028), metastatic disease (B = 16.2, p= 0.038), and outpatient opioid use (B = 32.8, p<0.001) were independent predictors of greater daily opioid administration during hospitalization. Conclusions: An evaluation of psychological distress and current opioid use may help to identify cancer patients at increased risk for more severe pain and higher analgesic requirements during hospitalization. This identification will facilitate the delivery of more intensive pharmacologic and nonpharmacologic interventions.