Computed tomography in patients with abdominal pain and diarrhoea: does the benefit outweigh the drawbacks?

The role of computed tomography (CT) in the evaluation of abdominal pain is well established. However, concern exists in regard to procedure-related radiation levels, contrast-medium toxicity and costs. We sought to determine whether the use of abdominal CT caused major changes in the management of patients who had abdominal pain and diarrhoea. We reviewed all abdominal CT examinations that were performed at our hospital from October through December 2010. We selected 574 scans that had been performed in patients who presented with or without diarrhoea. We examined the selected medical records to determine whether the CT scan changed patients' management. A scan was considered to be management changing if a decisive intervention occurred on the basis of the scan results. Among 124 scans in patients with diarrhoea and 450 scans in patients without diarrhoea, the scan results changed management in 13 of the patients with diarrhoea (11%) and in 233 of those without diarrhoea (52%) (P < 0.001). When management changed, the changes resulted from findings unrelated to diarrhoea. Despite its defined role in the assessment of abdominal pain, CT of patients that also had diarrhoea seldom caused a major change in management. The probability of CT causing such a change does not outweigh the cost, radiation risk or potential for contrast-induced nephropathy.

Recurrent histoplasmosis in AIDS mimicking a colonic carcinoma.

The prevalence rate of lower gastrointestinal bleeding in patients with AIDS is around 2.6%. A 42-year-old woman with AIDS (CD(4) count 9/microL) and recently treated for disseminated histoplasmosis presented to the emergency room with melena, severe anaemia and fever. A colonoscopy showed an umbilicated colonic nodule mimicking a carcinoma of the colon. The biopsy showed intracytoplasmic microorganisms compatible with Histoplasma capsulatum. She had poor compliance to the itraconazole when discharge on previous admission. Despite the fact that colonic histoplasmosis is uncommon, the mortality rate is around 8% and clinicians should be aware of the clinical presentation of histoplasmosis when recur, especially in patients not taking the itraconazole for long-term treatment.

Stenotrophomonas maltophilia pneumonia in cancer patients without traditional risk factors for infection, 1997-2004.

In order to elucidate the spectrum of Stenotrophomonas maltophilia pneumonia in cancer patients without traditional risk factors, 44 cancer patients (cases) with S. maltophilia pneumonia in whom S. maltophilia pneumonia risk factors were not present were compared with two S. maltophilia pneumonia risk groups (controls) including 43 neutropenic non-intensive care unit (ICU) and 21 non-neutropenic ICU patients. The case and control patients had similar demographic and underlying clinical characteristics. Compared with case patients with S. maltophilia pneumonia, neutropenic patients had higher exposure to carbapenem antibiotics (58 vs. 41%; p < 0.03), more frequent hematologic malignancy (95 vs. 64%; p < 0.0003), and they presented with concurrent bacteremia more often (23 vs. 0%; p < 0.0005). Patients with S. maltophilia pneumonia in the ICU needed vasopressor therapy more frequently than cases (62 vs. 5%; p < 0.0001). Hospital-acquired S. maltophilia pneumonia was more common among controls than cases (98 vs. 61%; p < 0.000002). Among the cases, 15 (34%) received outpatient oral antimicrobial therapy, while 29 were hospitalized and eight (28%) were subsequently admitted to the ICU. The mean duration of ICU stay, even among these eight patients (19 +/- 40 days), was comparable to that of patients with neutropenia (23 +/- 26 days) and those who developed S. maltophilia pneumonia during their ICU stay (34 +/- 22 days; p = 0.46). The overall infection-associated mortality in the 108 patients with S. maltophilia pneumonia was 25%. Twenty percent of patients without traditional risk factors for S. maltophilia pneumonia died due to progressive infection. In a multivariate logistic regression analysis, only admission to the ICU predicted death (odds ratio 33; 95% confidence interval, 4.51-241.2; p < 0.0006). The results of this study indicate S. maltophilia pneumonia is a serious infection even in non-neutropenic, non-ICU patients with cancer.