There have been several blogposts on this topic by colleagues in my sub-specialty, particularly on this blog. So, what could I possibly add? I just did a literature review on this topic for a talk and thought I would summarize this topic for my non-healthcare epidemiology colleagues and provide references for further reading. At the end, I will also share what I think we should do while waiting for that dream randomized controlled trial that is not going to happen.
Let’s go over some basics. Patients acquire new organisms in healthcare settings through contact with other patients with those organisms (patients touch other patients only rarely), by being touched by healthcare personnel with contaminated hands, by coming in contact with contaminated environmental items like devices and equipment, and by exposure to antimicrobial agents. For this blogpost, I want to focus on transmission via touch, and ways to prevent it in healthcare settings. It’s important to understand and appreciate the differences between standard precautions and contact isolation precautions. Standard precautions are to be used for every patient every time, and the precautions include hand hygiene, covering cough, use of appropriate personal protective equipment upon anticipation of exposure to blood or body fluids, injection safety, proper handling of medications and cleaning and disinfection of equipment. These are comprehensive, right? They are supposed to be part of our standard patient care work. The reality is that clinicians and healthcare organizations struggle to practice them successfully 100% of the time. Clinicians do not always correctly anticipate blood and body fluid exposure. That’s where contact isolation precautions come in. They specify the use of physical barriers, i.e., gown, gloves and any other prespecified personal protective equipment for every contact with the patient, not necessarily only when blood and body fluid exposure is anticipated. Unlike standard precautions that need to be used for every single patient every time, contact precautions (CP) are indicated for specific patients, e.g., those with multi-drug resistant organisms (MDRO) like methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococcus (VRE), extended-spectrum beta-lactamase producing Enterobacteriaceae, Clostridium difficile, or patients with draining abscesses or profuse diarrhea and incontinence.
So, how long do we need to keep a patient in contact isolation? The real answer is for as long as there is transmission potential justifying precautions beyond standard precautions. How long does this potential exist? We do not know. The reasons why decisions on duration are so complicated are the following.
- Lack of sufficient high-quality studies in non-outbreak situations
- CP alone are not enough to reduce transmission
- Transmissibility via touch depends on several factors – plasmid vs. chromosomal resistance, amount of pathogen (infection burden, colonization density), proportion of ‘dirty touches’, colonization pressure, and other factors
- Duration of colonization is hard to determine; Studies have shown results as wide as one to 200 weeks from time of colonization to clearance. Eradication of pathogen during the course of treatment is also hard to determine
- Technically, standard precautions may be sufficient in most cases. It’s complicated to study the effect of contact precautions vis a vis other infection prevention measures. That’s why we haven’t had that ‘magic’ randomized controlled trial addressing this question.
Why are more experts speaking out against contact precautions?
Contact precautions were initiated 50 years when infection control practices were very different from how we practice infection prevention now. Alcohol hand rub as we know it was introduced only around the year 2000. Several adverse side effects of contact precautions have been reported in recent years. Recent studies have found the following among patients in contact precautions compared to patients not placed in isolation.
- More frequent perceived problems with care upon patient HCAPHS survey (Mehrotra P et al. Infect Control Hosp Epidemiol 2013;34)
- More likely to complain to the hospital about care (8% vs. 1%); More likely to not have vital signs recorded (51% vs 31%); More likely to have days without physician progress note (26% vs 13%) (Stelfox HT et al. JAMA 2003;290)
- More likely associated with depression (Odds Ratio 1.4; 1.2-1.5) but not with anxiety (Day HR et al. Infect Control Hosp Epidemiol 2013;34)
- Delays in procedures, radiology studies (Morgan DJ et al. Infect Control Hosp Epidemiol 2015;36)
- Falls among MRSA/ VRE patients 4.57 vs. 2.04 per 1000 patient-days; Pressure ulcers among MRSA/ VRE patients 4.17 vs 1.19 per 1000 patient days. However, rates were not changed after removal of contact isolation. (Gandra S et al. J Hosp Infect 2014;88)
- VTE occurred in 17.7% of isolated patients and 3.5% of non-isolated trauma patients (Reed CR et al. J Trauma Acute Care Surg 2015;79)
- MRSA and VRE flags slowed down hospital operations with a mean time to bed arrival 9.63 vs 8.6 hours; mean length of stay 7.03 vs. 5.27 days; 1.19 (1.13-1.26) times more likely to experience an acuity-unrelated within-hospital transfer (Shenoy ES et al. Infect Control Hosp Epidemiol 2016; 37)
- Interns have been found to 2.3 vs. 2.5 visits per day, 2.2 vs. 2.8 min/ visit, 5.2 min/d per patient vs. 6.9 min/d per patient, All p-values <0.001 (Dashiell-Earp CN et al. JAMA Intern Med 2014; 174)
- More ≠ Better: Study of Compliance with CP in 11 Teaching Hospitals (Dhar S et al. Infect Control Hosp Epidemiol 2014; 35)
The reason experts are still divided is that these are single center observational studies and there is still a concern that MDRO rates might go up if contact precautions are decreased even if there is no evidence that they would. What are hospital infection prevention programs to do in the mean time? Make your best educated guess into policy and implement along with rigorous measures of effectiveness. Invest in improving compliance with standard precautions and improve culture of safety. Publish results. Teams led by Uslan DZ, Edmond MB, Almyroudis NG and Gandra S have done exactly that. Lastly, this viewpoint by Kathy Kirkland published in 2009 is as relevant now as it was back then.