Patients were identified from the Healthcare Utilization Project National Inpatient Sample (HCUP-NIS) for the years 1998-2002. The HCUP-NIS is a national, population-based sample representing 20% of hospital discharges annually in the United States and is prepared by the Agency for Healthcare Research and Quality. The number of states contributing data ranged from 22 states in 1998 to 35 states in 2002.
HCUP-NIS is an administrative database with information from a variety of hospitals that may differ in their coding practices. Detailed clinical data, such as symptom severity and fibroid size or location, cannot be described using the HCUP-NIS. Outpatient surgical procedures are not included in this database. Viagra Super Active
Patient Selection and Coding
International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes were used to select patients with leiomyoma of the uterus. Patients with a discharge diagnosis code of 218.xx (including any numbers to the right of the decimal point) or 654. lx (including any number for the x) were included. Subclassifications of ICD-9-CM code 218 specify the location of uterine fibroids, and 654.lx identifies uterine fibroids associated with pregnancy or labor and delivery. HCUP-NIS includes up to 15 diagnoses codes, and the first ICD-9-CM diagnosis code listed is considered the primary diagnosis.
To ensure a consistent uterine fibroid population, all patients with an admission source from the court or law enforcement (538 cases) and patients who had an admission from an “other source” (867 cases) were excluded. To ensure a comprehensive population, inclusion was not restricted by age; however, 153 cases were described as newborns and those subjects were excluded as were any subjects identified as male.
The types of procedures performed on the selected patient population are the primary focus of this study. In HCUP-NIS, as many as 15 four-digit ICD-9-CM procedure codes can be attributed to an inpatient visit. Currently, no ICD-9CM procedure codes are available for UAE, focused ultrasound, transvaginal cryomyolysis or myolysis.
Unadjusted means and percentages were calculated based on database counts using HCUP-NIS sample discharge weights. Multiple logistic regression analysis was used to estimate prevalence odds ratios for the association of the most common fibroid surgical treatments and patient and socioeconomic covariates of interest. The number of diagnosis and procedure codes as well as length of stay (LOS) were included in the analysis to account for potential differences in the severity of the condition and comorbidities. Reference categories were identified for relevant variables. The detailed definitions of these variables are further elaborated in HCUP-NIS documentation.
Statistical Package for the Social Sciences (SPSS) version 13 was used for all analyses.