Abstract
The Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH) was a randomized double-blind placebo-controlled trial to test whether hydroxyurea could reduce the rate of painful crises in adult patients who had at least three painful crises per year. Daily pain diaries were collected biweekly for 299 MSH patients. Patients rated daily pain on a linear scale 0 through 9. The daily pain data was collected to provide useful information regarding the efficacy of hydroxyurea in reducing the rate of painful crises. The objective of this analysis was to determine if there was an underlying time series pattern associated with the MSH pain diary data, especially one indicative of a calendar or seasonal association. An Unobserved Component Model (UCM) was used to determine, based on empirical data, if there was a cycle and/or trend associated with the time series. The times series started on December 1991 and ended on December 1994. The series was equally spaced and univariate. The equal spacing was a monthly interval. The average pain score for each two- week diary period was used to create each unit of analysis. This analysis considered three groupings of patients: all patients, hydroxyurea patients and placebo patients. The results of the modeling indicated that there was indeed both a cycle and a trend associated with the MSH pain diary data. This is the case for each group of patients considered. All three groups had a cycle pattern that seemed to show a seasonal behavior of average pain scores (the changing of the season seemed to show an approximate transition from trough (Spring) to peak (Fall/Winter) and peak to trough). There was a downward trend in average pain scores for both the all patientsgroup and the hydroxyurea patientsgroup. However, there was an upward trend in average pain score for the placebo patientsgroup. This may reflect the result shown in the MSH Clinical Trial, that HU usage reduced the average pain scores reported by patients.
Disclosure: No relevant conflicts of interest to declare.
Author notes
Corresponding author
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal