Low Bacterial Diet in Patients with Cytopenia after Intensive Chemotherapy for Hematological Malignancy: A Study of Efficacy and Costs.

Patients with hematological malignancies who receive intensive chemotherapy usually develop a period of cytopenia, during which there is an increased risk of infection. In order to reduce the risk of infection several preventive measures have been adopted. Fundamentally, all of these measures were designed to prevent either acquisition of Gram negative rods or fungal pathogens from the environment, or the translocation of these potential pathogens across the mucosal barrier of the gut. The evidence for the necessity of low bacterial diet as an infection preventive measure is weak. Therefore, in this randomized, controlled study the question whether normal hospital diet is equivalent in efficacy to low bacterial diets, given as an infection preventive measure during treatment of cytopenic patients, was addressed. Intestinal colonization by aerobic Gram negative rods and yeasts has been chosen as primary endpoint. In addition, the occurrence of infections and the total costs of hospital care have been documented, in order to identify potential cost savings by the use of either diet. The patients were randomized into two groups: one group (Study group) to receive antimicrobial prophylaxis (AP) and low-bacterial diet, the other (Control group) to receive the same AP and normal hospital diet. The primary outcome parameter is colonization of faeces with Gram-negative rods or Candida species. Degree of colonization was calculated as [∑ 10log (CFU/g feces)] x [duration of episode] /[number of days on which feces culture were taken]. Next, an analysis of variance was performed taking several factors into account, such as the use of antibiotics. As secondary outcome parameters all infections, and therapeutic use of antibiotics were registered; moreover, total societal costs were calculated for each patient starting at first hospitalization until 28 days after the last discharge. In the study group 10 patients were included and evaluable, in the control group 10 patients were included and 9 evaluable. The number of cycles studied, defined as the period from starting antibiotic prophylaxis until leukocyte counts had recovered to 1000 /mm3 or higher, was 21 in the study group, and 18 in the control group. The median number of days per cycle was 28 (range 17–43) days for the study group, and 34 (range 19–60) days for the control group (NS). Regarding the primary outcome parameter, gut colonization by yeasts or Gram negative rods, no statistically significant differences between treatment groups were observed. With respect to infections as secondary outcome parameter, neither the number of days with fever nor the number of days, during which antibiotics were administered therapeutically, was different between the study and control group of patients. The numbers of fungal infections were no cases of invasive aspergillosis (IA) in the study group, 2 episodes of possible aspergillosis in the control group, and in both groups one episode of candidemia. Regarding the total societal costs no statistically significant differences were measured between normal diet and low-bacterial diet. The results indicate that low-bacterial diet and normal hospital diet are equivalent regarding colonization with yeasts and Gram negative rods, infections and total societal costs.