Schlesinger, A., Paul, M., Gafter-Gvili, A., Rubinovitch, B., & Leibovici, L. (2009). Infection-control interventions for cancer patients after chemotherapy: a systematic review and meta-analysis. The Lancet Infectious Diseases, 9, 97–107.

To review prospective comparative studies that address infection prevention for high-risk patients undergoing chemotherapy and stem cell transplant recipients to evaluate the evidence for best practice and assess for mortality.

Databases searched were CENTRAL (the Cochrane Library issue 4, 2006), PubMed (1966–2008), and LILACS (1982–2006). Unpublished trials that were presented at the following conferences were also queried: American Society of Hematology (2001–2006), American Society of Clinical Oncology (1995–2006), and European Society for Medical Oncology (2002–2006). The references lists from all included studies were reviewed to identify additional studies. 

Keywords searched were not provided by the authors.

Studies were included if they 

• Were prospective comparative studies that included patients with cancer in the hospital or outpatients who were receiving chemotherapy for solid tumors.
• Reported hematological malignancies.
• Reported stem cell transplant recipients.  
• Compared an intervention to placebo, no treatment, or another intervention, as well as all environmental measures, barrier precautions, and other nonpharmacological measures used for the prevention of infectious diseases.

Studies were excluded if they were nonrandomized studies that compared patients with different types of cancer or compared different treatment protocols (i.e. stem cell transplant versus chemotherapy) or if they assessed pharmacological interventions, such as antimicrobial prophylaxis and mouth rinse preparations, unless these interventions were applied together or as a control for the infection-control interventions.

Forty studies were included.

Method of Study Evaluation

Prevention of infection measures used in the studies were identified and grouped together: control of air quality (air filtration), protective isolation, and suppression of the endogenous flora (antibiotics). Subgroup analyses were performed for two major outcomes: all-cause mortality at 100 days and documented infections.

• The total sample size was 40 studies; the sample range across studies was 45 to 3900.
• Twenty-six studies in the analysis studied protective isolation, 11 assessed outpatient versus inpatient care, and three evaluated other random interventions.  
• Twenty-nine studies assessed patients with acute leukemia, six included patients with other hematologic malignancies, and 22 included hematopoietic stem cell transplant (HSCT) recipients exclusively.
• No study included patients with solid tumors at low risk for infection.

The primary outcome, all-cause mortality, was assessed at 30 days, 100 days, and at the longest follow-up reported in each study. Secondary outcomes included the rate and types of infections (including bacteremia), the need for hospitalization and length of hospital stay, the length of the febrile period, infection-related mortality, bacterial and fungal colonization, and antibiotic and antifungal treatment.

• Protective isolation, including control of air quality, barrier isolation, and endogenous suppression due to antibiotics, were shown to significantly reduce mortality (relative risk [RR] = 0.79; 95% confidence interval [CI] [0.72, 0.87]).
• Antibiotic prophylaxis was the main component mediating the beneficial effect of isolation interventions. Improved survival was only shown when antibiotic and antifungal prophylaxis was used with air quality control or barrier isolation (RR = 0.66; 95% CI [0.55, 0.79] in studies with prophylaxis compared with RR = 0.93; 95% CI [0.75, 1.15] in studies without prophylaxis).
• Protective isolation with prophylactic antibiotics also significantly reduced infection, whereas control of air quality or barrier isolation alone did not.
• Control of air quality did not significantly reduce mold infections.  Outpatient mortality was also significantly lower than inpatient mortality (RR = 0.72; 95% CI [0.55, 0.95]).

A combination of air quality control, barrier isolation, and prophylactic antibiotics is estimated to reduce 30-day all-cause mortality by 40% in high-risk cancer patients, including allogeneic or autologous HSCT recipients and patients with acute leukemia. Prophylactic antibiotic use is the most significant preventive strategy; air quality control and barrier isolation did not show an independent contribution, and their use should be reserved for patients at highest risk of infection. Survival significantly improved with outpatient management, which included the use of prophylactic antibiotics without environmental controls.

The reduced mortality seen in outpatient management may be due to

• Selection bias (healthier patients are able to remain outpatient)
• Decreased risk of infection from hospital pathogens
• Use of antibiotic prophylaxis.

There is no evidence that any of these preventive strategies (air quality control, barrier isolation, or prophylactic antibiotics) are necessary or appropriate for low-risk cancer patients with solid tumors.