Constipation is a prevalent comorbidity affecting ∼50% of patients with long- term opioid therapy. In clinical routine different diagnostic instruments are in use to identify patients under risk. The aim of this study was to assess the diagnostic performance of an 11-item Likert scale for constipation used as a self-assessment in opioid-treated patients. This trial was conducted as a retrospective cohort study in Berlin, Germany. Patients with long-term opioid therapy treated in 2 university-affiliated outpatient pain facilities at the Charité hospital were included from January 2013 to August 2013. Constipation was rated in a self-assessment using a numeric rating scale from 0 to 10 (Con- NRS) and compared with results from a structured assessment based on ROME-III criteria. Altogether, 171 patients were included. Incidence of constipation was 49% of patients. The receiver-operating characteristic of Con-NRS achieved an area under the curve of 0.814 (AUC 95% confidence interval 0.748–0.880, P < 0.001). Con-NRS ≥ 1 achieved sensitivity and specificity of 79.7% and 77.2%, respectively. The positive predictive value and the negative predictive value were 70.3% and 81.6%, respectively. Overall diagnostic performance of a concise 11-item Likert scale for constipation was moderate. Although patients with long-term opioid therapy are familiar with numeric rating scales, a significant number of patients with constipation were not identified. The instrument may be additionally useful to facilitate individualized therapeutic decision making and to control therapeutic success when measured repetitively.
