Douglas V. Sepkowitz

Spinal Epidural Abscess in Hemodialysis Patients: A Case Series and Review

Spinal epidural abscess (SEA) is a rare but potentially devastating condition. We noticed an increase in the number of cases of SEA, with the majority in hemodialysis (HD) patients. This prompted a retrospective chart review of all cases of SEA admitted from 2000 to 2005 and a literature search of similar cases. We identified 19 SEA cases treated at Long Island College Hospital during this 6-year period, of which six were on HD: four were dialyzed via catheter, one via arteriovenous fistula, and in one the method of dialysis was not documented. Four patients had bacteremia with Staphylococcus aureus. Four patients presented with paresis or paralysis; only one improved. The mortality rate was 33% (2/6). We found 30 other cases of SEA in patients on HD from the literature. These 36 HD cases were compared with 85 SEA cases that were not on HD (13 from our study and 72 described in two large case series). The mortality rate was noted to be much higher in HD patients (23% [6/26] versus 7% [6/85]). Neurologic deficit at presentation was noted in 47% (17/36) of HD patients versus 69% (59/85) of non-HD patients, but neurologic improvement was higher in non-HD patients (71% [42/59] versus 29% [5/17]). This is the largest literature review of SEA in patients on HD. When compared with non-HD patients, HD patients had a higher mortality rate and were less likely to improve neurologically.

Massive Hepatic Enlargement with Fatty Change Associated with Ketoconazole

Several forms of hepatic toxicity have been described with the antifungal agent ketoconazole. We report a case of massive liver enlargement with fatty infiltration presenting as gastric compression. This occurred in a young woman with AIDS taking ketoconazole as maintenance therapy for cryptococcal meningitis. This is the first reported case of ketoconazole causing fatty change in the liver.

Four Cases of Spinal Tuberculosis Within 7 Months at a Community-Based Hospital

AbstractSpinal tuberculosis is rare in the United States. Among the 914 tuberculosis cases in New York city in 2007, there were only 19 cases (2%) involving bone/joint as the primary site. It was surprising when there were 4 new cases within 7 months in our hospital in Brooklyn, New York City.

Sa1237 Epidemiological Features and Risk Factors Affecting Outcomes of Clostridium difficile Infection Among Inpatients At an Urban Hospital

G A A b st ra ct s in an inpatient cohort from Olmsted County, MN. Information on maximum change in serum creatinine (highest value compared to baseline over the past year) and WBC within 7 days of CDI diagnosis was obtained for three groups: pediatrics ( ,18 yrs old), adults (18-65), and the elderly (.65). Sensitivity (Se) and specificity (Sp) of various levels of WBC and creatinine ratio were estimated for predicting severe-complicated CDI, defined as need for ICU admission, colectomy, or death. Analysis of creatinine ratio in the pediatric population was not done due to a high proportion of missing baseline creatinine values. The cohort included 1446 patients (48.6% female, median age 62.5 yrs [range, 0.1-103.7]); 487 (33.7%) had severe-complicated CDI. A WBC .15 x10^9/L (the cut-off recommended in expert guidelines) had a similar Se and Sp for predicting severe-complicated CDI in pediatrics (Se= 54.5, Sp=78.8), adults (Se=53.7, Sp=73.2), and the elderly (Se=56.3, Sp=63.5). A serum creatinine ratio of 1.5 had a similar ability to predict severe-complicated CDI in adults (Se= 59.2, Sp=59.5) and the elderly (Se=61.9, Sp=54.1). Lowering the thresholds of WBC and creatinine ratio increased the Se but decreased the Sp for predicting severe-complicated CDI in all age groups. For example, lowering the WBC threshold to 12.5 or decreasing the creatinine threshold to 1.4 increases the Se by 10-18% with a similar decrease in Sp (Table 1). No optimal WBC or creatinine rise cut-offs to predict severe-complicated CDI were defined in this investigation. The current cut-off values according to expert guidelines performed similarly across age groups, but had relatively low sensitivities and specificities. For all age groups and for both WBC and creatinine rise, decreasing the cut-off value increased the sensitivity while decreasing the specificity; these variables should be considered as continuous predictors of complicated CDI, with higher values indicating higher risk. Values below the threshold values defined in current guidelines do not reliably exclude a significant risk of complicated CDI. Given the poor outcomes in patients with severecomplicated CDI and the recommendation that treatment be stratified by severity, a more sensitive predictor (lower WBC or creatinine cut-offs) may be more appropriate. Table 1. Sensitivities and specificities of peripheral leukocyte count and serum creatinine rise in predicting severe-complicated Clostridium difficile infection.