Health Care-Associated Infection (HAI) Rates Are More Pertinent Quality Indicators Than Nosocomial Infection (NI) Ones for Hematology Programs.

Sami Chehata, Chiraz Grira, Patrick Legrand, Cecile Pautas, Sebastien Maury, Mathieu Kuentz, Jean Carlet, Catherine Cordonnier

Author address: 

Hematology, Henri Mondor Hospital, Creteil, France; Hematology, Institut Gustave Roussy, Villejuif, France; Microbiology, Henri Mondor Hospital, Creteil, France; Intensive Care Unit, Hopital Saint-Joseph, Paris, France


In hematology patients, HAI are a complex mixture of infections due to therapy, especially to transplant procedures, conditioning regimen, immunosuppressive drugs, and invasive procedures. These HAI may occur within the hospital and be related to the hospital environment (i.e., NI), or not. These different kinds of infections require very different preventive measures. The objective of our study was to assess the incidence, overlap, and causes of HAI and NI in our hematology ward over a 6 month period. Methods and Definitions: Infections were considered to be HAI, irrespectively of their site of occurrence (hospital or not) when they were directly related to the therapeutic procedures and subsequent immune deficiency. This included: febrile neutropenia (FN) subsequent to chemotherapy, catheter-related infections, herpes virus reactivations in patients previously sero+ to the virus, primary CMV infection in CMV- patients transplanted with a CMV + donor, toxoplasmosis reactivation, invasive fungal infections, and any infection occurring within 6 months of transplant, or later in case of immunosuppressive treatment after this date. Infections were considered to be NI when they occurred > 48 h after hospitalization without any symptom of infection at hospitalization. Aspergillosis was considered to be NI when there was a delay of 7 days between hospitalization and first symptoms.We retrospectively assessed 223 consecutive infectious episodes occurring between 7-12/2000 in 137 patients, including 59 stem cell transplant recipients (allogeneic: 23; autologous: 36). Results : 204/223 infectious episodes (91,4%) were HAI while 94/223 (42%) were NI. 7/223 (3,6%) were causes of death. Among the 42 bacterial documented infections, 20 (47.6%) were due to potentially hospital-acquired strains: 16/23 (69.5%) of HAI-NIs, and 4/19 (21%) of the non NI-HAIs. Conclusion: In a preventive infection program, the assessment of NI clearly underestimates half of the HAI in hematology patients. Although hospital environment control is of paramount importance in transplant programs, we suggest that HAI should be a better quality indicator for hematology programs than NI.

abstract No: 


Full conference title: 

47th American Society for Haematology
    • ASH 47th (2005)