How much does the type of tympanostomy tube matter? A utility-based Markov decision analysis.

Abstract

To use a utility-based Markov decision analysis model to compare outcomes of short-term grommet tympanostomy tubes (TTs), intermediate-type tubes, and permanent T-Tubes and to use a detailed sensitivity analysis to determine the most important factors influencing outcomes with one type of TT versus another. Markov decision analysis. Hypothetical cohort. A Markov cohort decision analysis model was created using computer software (TreeAge Software, Inc, Williamstown, Massachusetts) comparing the 3 types of TT. Published data were used to determine key baseline model parameters. The model featured potential complications including eardrum perforation, early TT extrusion/blockage, and the need for possible repeat TT placement after extrusion. Outcomes were quantified using a 0.95 (1 procedure, full recovery) to 0.6 (failed myringoplasty) utility scale. Utility values were discounted over time to incorporate real-life inconvenience. The intermediate TT accumulated superior total utility in 2-, 4-, and 6-year models (2.48, 3.96, and 5.27 total utility) compared with grommets (2.32, 3.82, 5.18) and T-Tubes (2.42, 3.86, 5.18). Examining a hypothetically otherwise healthy 3000-child cohort, T-Tubes resulted in an increased overall persistent perforation rate (7.9% vs 0.2%, P < .001, χ(2)). Sensitivity analysis indicated that the rate of repeat TT placement, the utility assigned to having a functional tube in place, and the inconvenience utility discount factor over time were the driving factors of the model. The intermediate TT may produce optimal outcomes as it combines a balance of a lower perforation rate than T-Tubes and a longer period of ventilation than grommet tubes.