ABSTRACT
OBJECTIVES: To identify correlates of a prolonged length of stay (PLOS) in women hospitalized for preeclampsia/eclampsia in Texas, USA. METHODS: Statewide hospital data were obtained, and the records of women who were discharged in 2004 and/or 2005 with a principal discharge diagnosis of preeclampsia or eclampsia were extracted using ICD-9-CM codes. PLOS was defined as a stay greater than 5 days. Odds ratios (OR) for PLOS were calculated. Generalized estimating equations were used to account for a small group of women who were hospitalized multiple times during the study period for preeclampsia. A total of 21,203 records were analyzed. RESULTS: The crude incidence of PLOS was 17.5%. Advancing maternal age was positively associated with PLOS: for every 10-year increase, there was a 20% increase in the odds of PLOS (adjusted OR = 1.20,95% confidence interval (CI): 1.13, 1.28). The strongest risk factor for PLOS was the presence of renal disease: adjusted OR 5.81 (95% CI: 3.97, 8.50). Protective factors included Medicaid beneficiary status, and being admitted from the emergency department. CONCLUSIONS: The strongest correlate of PLOS in a large cohort of women hospitalized for preeclampsia was the presence of renal disease.
Subject(s)
Eclampsia/epidemiology , Kidney Diseases/epidemiology , Pre-Eclampsia/epidemiology , Comorbidity , Confidence Intervals , Female , Humans , International Classification of Diseases , Length of Stay , Maternal Age , Medicaid , Odds Ratio , Patient Discharge , Pregnancy , Risk Assessment , Risk Factors , Severity of Illness Index , Texas , United StatesABSTRACT
OBJECTIVE: To determine if the established endometrial thickness cut point (5 mm) for abnormal endometrial pathology shifts to higher thickness in the presence of selected risk factors/comorbidities. STUDY DESIGN: A sample of 112 postmenopausal women was identified. The outcome was abnormal endometrial pathology, be it endometrial cancer or hyperplasia with atypia. Logistic regression was used to calculate prevalence odds ratios (ORs) of abnormal results for women with thick or thin endometria and 0 or > or = 1 of the following comorbidities/cofactors: obesity, diabetes, hypertension and use of hormone replacement therapy. RESULTS: Approximately half the sample was hypertensive; 56.3% were obese. A large proportion (84.8%) of the patients had > or = 1 of the comorbidities/cofactors of interest. Women with endometria > or = 12 mm and > or = 1 comorbidities appeared to have 5 times the odds of having an abnormal result compared to women with thin endometria (<12 mm) who had 0 comorbidities; this result was not statistically significant (adjusted OR = 5.08, p = 0.07). A dose-response curve (regression spline) showed that the prevalence of an abnormal outcome increased sharply between 5 and 9 mm. CONCLUSION: Clinicians should continue to use the 5-mm cut point when deciding whether patients should have endometrial sampling.
