No abstract

Children experience moderate to severe pain in the days after tonsillectomy surgery. This article describes the challenges of analyzing data from pain diaries 24 hours after ambulatory tonsillectomy surgery. Instructions were to record pain levels every 4 hours and the analgesic as given; however, the number of entries was inconsistent, making comparison of groups difficult. The use of analgesics can threaten the construct validity of cause and must be considered and controlled statistically. Opioids were converted to morphine equivalents and the nonopioids were held as a separate covariate. This article describes the steps taken to find an appropriate statistical means of dealing with inconsistent diary entries and analgesic use. Hierarchical linear modeling (HLM) provides a means to analyze unbalanced data and control for analgesics. Issues of control must be addressed in research that measures children's pain in the home. HLM can address issues of incomplete data in repeated-measures studies.