Patient and physician agreement on abdominal pain severity and need for opioid analgesia

Am J Emerg Med. 1999 Oct;17(6):586-90. doi: 10.1016/s0735-6757(99)90203-6.

Abstract

Whereas controversy surrounds emergency department (ED) analgesia administration to patients with undifferentiated abdominal pain, few studies have addressed the level of patient-physician agreement on abdominal pain severity and need for opioid analgesia. This prospective study was undertaken to assess concordance between emergency physicians and patients on abdominal pain severity. Study subjects were a convenience sample of 30 adults seen in an urban university-affiliated tertiary care ED (annual census 65,000) who had undifferentiated abdominal pain meeting an initial severity threshold of 5 on a 10 cm visual analog scale (VAS) marked by the patient. Patients' and physicians' VAS scores, obtained in blinded fashion at presentation (t0) and at one (t1) and two (t2) hours into the ED stay, were compared with t test (VAS scores) and sign-rank (percent change in VAS scores) analyses. In addition, patients and physicians were asked at each assessment time, in blinded fashion, "Is the pain severe enough to warrant morphine?" The kappa statistic was used to characterize the degree of agreement between physician and patient assessments as to whether opioids were indicated. At t0, t1, and t2, patients' mean VAS scores (7.5, 6.7, and 5.1) were significantly (P < .05) higher than the corresponding physicians' VAS scores (5.3, 4.7, and 3.9). Though VAS scores for physicians started lower than those of patients, the percentage changes in scores from one assessment to the next were similar by Wilcoxon sign-rank testing (P > .50 for time intervals t0 - t1 and t1 - t2). Overall, patients and physicians agreed on the question of whether pain was sufficient to warrant opioids in 71 of 90 (78.9%) assessments; the corresponding kappa statistic of .57 indicated moderate agreement (P < .0001). These results, indicating that patients and physicians usually agree on whether opioids are warranted for abdominal pain, have important implications for further research on ED analgesia in this population.

MeSH terms

  • Abdominal Pain / diagnosis*
  • Abdominal Pain / drug therapy*
  • Adult
  • Analgesics, Opioid / therapeutic use*
  • Emergencies
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Observer Variation
  • Pain Measurement / methods*
  • Prospective Studies
  • Self-Assessment
  • Statistics, Nonparametric

Substances

  • Analgesics, Opioid