Commentary on Variations in Health Insurance Policies Regarding Biological Therapy Use in IBD and a Modified Delphi Process for the Development of Choosing Wisely for IBD
Nguyen et al rigorously applied the modified Delphi process to select 5 recommendations for the Choosing Wisely list for IBD. The Choosing Wisely initiative, sponsored by the ABIM foundation, challenges medical specialty societies to identify commonly used tests, procedures, or practices that may be unnecessary, harmful, duplicative, or not supported by evidence. The 5 recommendations for IBD care included (1) not using steroids for maintenance therapy, (2) avoiding the use of opioids for long-term pain management, (3) avoiding the use of intravenous corticosteroids alone for more than 5 days in patients with acute severe ulcerative colitis, (4) not relying on symptoms alone when evaluating disease activity, and (5) avoiding the use of abdominal computed tomography (CT) in the acute setting unless one suspects an IBD-related complication or a non-IBD cause for symptoms. Although all these recommendations help to identify interventions that are potentially unnecessary or harmful, the recommendation regarding CT use is challenging to apply when faced the reality of caring for this patient population.
Several groups have attempted to develop prediction models with the goal of reducing unnecessary CT use in emergency departments (EDs). The first study included 2 EDs in one academic health system and developed a risk score to predict an outcome of perforation, obstruction, abscess, or non–Crohn's disease (CD)-related urgent findings (APON). A risk score was developed which subtracted 1 point for hematochezia and added 1 point for history of obstruction, history of abscess, leukocyte count >12,000/μL and determined that among patients with a risk score of −1, the predicted risk for APON was 9%.1 A study in 11 Korean EDs using the same outcome identified independent positive predictors of structuring or penetrating disease (odds ratio [OR] 2.72, 95% confidence interval [CI] 1.21–6.13), leukocyte count >10,000/mm (OR 4.38, 95% CI 2.10–9.13), heart rate >100 beats/min (OR 2.33, 95% CI 1.10–4.93), and C-reactive protein >2.5 mg/dL (OR 3.11, 95% CI 1.23–7.86).2 Finally, Govani et al3 developed a clinical prediction model in a population of patients with CD in a single tertiary care ED to predict perforation, abscess, or other clinically significant findings. They found that a model using C-reactive protein and erythrocyte sedimentation rate alone with a predicted probability cutoff of 6% had a sensitivity of 93.8% and a negative predictive value of 98.1%. Unlike the other models, obstruction was not included in the outcome as no model demonstrated good performance characteristics to predict this outcome. Although these 3 studies are important contributions, they should not be used to guide clinical decision making. Prospectively, validated models are needed using an unselected population of patients with CD presenting to the ED with abdominal pain.
The study by Yadav et al published in this issue demonstrates the discrepancy between insurance policy criteria for approval of biological therapy and the AGA guidelines regarding the use of biologics in patients with IBD. They reviewed the policies available on-line for 50 of the top 125 insurance companies by market share in 2014 and found that 98% of policies were inconsistent with the AGA ulcerative colitis (UC) pathway,4 and 90% of the policies were inconsistent with the AGA CD pathway.