Sitting and Watching People - Thoughts on Lack of Standard Practice in Infection Prevention

Guest blogger:  Dr. David Hartley, PhD MPH is a Research Associate Professor in the department of Microbiology and Immunology at the Georgetown University Medical Center. His research interests include the ecology of infectious disease, public health surveillance, hospital infection control, and biological defense. His research applies mathematical modeling methodologies to understand the dynamics of disease in human and animal populations.

Patients who undergo hematopoietic stem cell transplantation are often profoundly immuno-incompetent for months post-transplant. Individuals typically neutrophyl engraft within weeks of the procedure, but in the case of imperfectly matched transplants, complications such as GVHD can make ongoing immunosuppressive therapy necessary in order to achieve t-cell tolerization. During this period, there is potential for infection -- newly acquired and recrudescence, both with attendant potential for emergent drug resistance. The medical management of such individuals can be complex and often results in admissions and tests/procedures in multiple hospitals and departments. In between hospital stays, such patients typically receive long term care in their homes.

As a non-MD epidemiologist who has observed infection control as it's actually practiced in relation to this patient population, I'm left with several impressions:

First, at the individual level, I am struck by the variation between healthcare workers in terms of infection control technique and compliance. Some caregivers can't seem to be bothered to tie their gowns, and I've seen more than a few put the gowns on only up to their elbows. With the gown literally hanging off their forearms, perhaps they don't realize the opportunity for clothes to contaminate or become contaminated with infectious material as they interact patients. Regarding hand washing, the length of soaping and rinsing can range from a few seconds to the time it takes to sing several stanzas of "happy birthday to you".

Second, at the department level, I've noted tremendous heterogeneity in infection control practice between wards within a given institution. The differences between CVL usage and care in post-op versus hematology/oncology units can be stark, not to mention differences in awareness of the importance placed on hand hygiene and isolation of immunosuppressed patients from others.

Third, there are differences in policy and practice between institutions. At some facilities, for example, when indicated, family caregivers and patient advocates must employ contact precautions (i.e., wear gowns and gloves) only when they are in patient rooms, while at other facilities such individuals are required to wear gowns and gloves only when they leave the patient's room (e.g., while going to restrooms or conference/consult areas). In some places the donning or shedding of such garb takes place inside the patient's room, while in other places it happens outside the room.

Lastly, there may be important interactions between the home and hospital environments. Heme-onc and transplant floors are typically controlled, well engineered areas in terms of air handling, environmental cleaning, and visitor access. Home environments, on the other hand, can span many logs of microbial contamination as a function of season, age and type of home, presence of pets, maintenance of HVAC systems, and other factors. Having patients transiting back and forth between these environments further complicates the epidemiologic picture.

This critically fragile patient population needs effective infection control practice in and out of the hospital, better surveillance, and careful epidemiological study. How are we to understand the differences in infection rates between wards or institutions given these and other inhomogeneities in practice? How can studies evaluating practice or intervention be valid when there is so much "noise" in the system? We need evidence-based guidelines and verification of compliance.