Annual antibiotic resistant infections in the U.S.

Just as the CDC's and other experts' thoughts on Ebola and infection control have evolved with experience, mine have taken a slight twist as well. Given the missteps at Dallas’s Texas Presbyterian Hospital, which could have occurred in any community, I now agree with the current recommendations to centralize care in specialty centers. My perspective has changed a bit, particularly given more information about the voluminous diarrhea and vomiting of Ebola victims, and the recent infections of health care workers.

But I feel a sadness, too, that by only having special Ebola SWAT teams trained in response, we are losing a critical and rare opportunity to improve infection control training for all.

We know there is an urgent and neglected need to control superbugs. “Each year in the United States, at least 2 million people become infected with bacteria that are resistant to antibiotics and at least 23,000 people die each year as a direct result of these infections. Many—if not most—of these would be preventable.

Impact of C. diff infections in U.S. annually.

For example, Clostridium difficile (aka C. Diff) causes 250,000 infections per year and 14,000 deaths, as well as $1,000,000,000 in excess medical costs per year.

Infections from carbapenem-resistant Enterobacteriaceae (CRE) bacteria are rapidly rising, and these bacteria are resistant to almost all antibiotics. According to the CDC, more than 9,000 healthcare-associated infections are caused by CRE each year, with 600 deaths.

Impact of CRE in US annually

In recent posts ("Superbugs" and "Public Health") I stressed this and emphasized the need for training and retraining all HCW and support staff in proper infection control techniques.

This was further highlighted this week by the tragic story of Troy Stulen,the 20 year old who survived a bone marrow transplant only to die at NIH of an infection that was resistant to all antibiotics, a carbapenem-resistant organism (CRE). Unfortunately, hospital administrators will not spend the money to train or retrain their staff without some serious incentives, and the focus now, though it makes sense, to have Ebola response teams just took the heat off for more general improvements in infection control.

CDC’s new recommendations

I heartily agree with the main elements of the CDC statement:

— First, they will make sure that health care workers dealing with Ebola patients are "repeatedly trained," especially when it comes to learning how to put on and take off their personal protective equipment.

— Second, the equipment used should leave no skin exposed.

— Third, these regulations should be monitored by a "trained observer" or site manager, who watches each employee take on and off their personal protective equipment

The emphasis on the need to practice and have a buddy to observe and help are especially welcome recommendations. Again, these precautions should include EMTs, housekeeping, radiology techs, etc.—everyone who might come in contact with infectious patients or their secretions.

Here’s where the CDC and I differ, or this week’s (10/20) press conference was unclear to me.*

Evaluation of Returned Traveler for Ebola; click to enlarge. (Source: CDC)

The 10/2/14 CDC algorithm for evaluating risk is an excellent infographic except for not addressing patients who will present with fever and flu-like, upper respiratory tract symptoms. It is appropriate and helpful for clinics or emergency room settings, saying:

1. Isolate patient in single room with a private bathroom and with the door to hallway closed

2. Implement standard, contact, and droplet precautions

The new CDC recommendations, however, suggest that for any suspected Ebola patient, which could be interpreted as anyone with the appropriate travel history and fever or “symptoms,” an Ebola SWAT team would be called, and that appropriate attire would be the “The Works”—aka the “Full Monty”—with Hazmat suits and N-95 masks or Powered Air Purifying Respirators (PAPRS). The proposed Hazmat PPE makes sense for patients with vomiting and diarrhea and for ICU patients, who presumably have a higher viral load and are more infectious.

As an infectious disease physician in community hospitals, I believe the CDC has made a slight misstep in its new recommendations. I would favor a tiered approach to a response, as follows.

While I understand the need for an Either/Or, all-or-none (go/no-go) decision point in an algorithm, this doesn’t make sense for patients who are not very ill. From a community hospital perspective, this is neither practical nor likely necessary, and will engender more fear and panic, furthering the current epidemic of Ebolanoia. We need to emphasize that most patients in Kikwit and elsewhere have been handled safely without hazmat suits. And even in Dallas, where many people were exposed to Thomas Duncan before he was properly identified as having Ebola and isolated, only the two ICU nurses became infected.

Instead, I would recommend the following tiered approach (details below):

Proposed Risk-based Ebola isolation algorithm

If Ebola might be suspected because of travel or exposure history, initially isolate the patient using standard, contact and droplet precautions.

Flu vs. Ebola

If they have fever and respiratory tract symptoms, as is likely during flu season, use the Personal Protective Equipment (PPE) as per the previous CDC precautions for multi-drug resistant organisms: an impervious gown, gloves, but enhanced with a mask (surgical or N-95) and face shield. Ebola is possible, though less likely. This would be akin to the CDC algorithm’s low-risk exposure.

If the patient presents with GI symptoms sycg as vomiting and diarrhea, the likely infectious risk is higher (whether norovirus or Ebola), so either an impervious gown plus leg coverage or a Tyvek hazmat suit makes better sense. In speaking with several epidemiologists, the Tyvek suits and PAPRs are considerably more comfortable, especially for the nurses who are in contact with the patients for much longer stretches.

If Ebola is confirmed or there was a high-risk exposure and the patient is symptomatic, then it makes sense to have the Ebola SWAT response team activated. It is just too easy for someone inexperienced to contaminate themselves removing PPE, which has been shown previously.

Practicalities in new Infection Control

My personal perspective on isolation attire:

Gloves: Vinyl gloves are often used in nursing homes and in some hospitals, as they are less expensive than latex. They also tend to be ill-fitting and do not adequately cover at the wrist, leaving skin exposed. They would never be used with an Ebola patient. Double-gloving makes sense. More than that would likely make it cumbersome to work and might increase errors.

Garbed for Superbug Isolation Patient

Gowns: For low-medium risk exposures with any isolated patient, especially those I see with multi-drug resistant organisms (MDROs), I prefer the vinyl isolation gowns to the cloth ones, because I can tear these off without possibly contaminating myself while untying strings at the waist or neck. There is no risk of the sleeve sliding up over the glove, leaving skin exposed at the wrist. I also like the vinyl gowns because, when removing gloves, I don’t have to slide my finger under the glove—I can just peel everything off at once. **A video from Emory University shows an excellent way of removing gloves with a "beaked" gloved hand (at 1:30).

Removing gloves - cdc

Masks: Surgical masks are fine for droplet secretions, and we have no reason to believe that Ebola has airborne transmission. N-95s are difficult to fit and uncomfortable to wear for even a few minutes. I have only used one type of PAPR, and it made communication and examination rather difficult, but was far more comfortable than the N-95. For low-intermediate risk settings, if a patient is not spewing infectious fluids, I disagree with Dr. Frieden’s statement: “we are recommending either of those options, but not a face mask and that's not because we think that Ebola is airborne, but rather because we think that what gets done in American hospitals can be so risky, whether that's suctioning or intubation or other things that may not be done in other parts of the world or in Africa.” This is surely true in the ICU, but seems overkill otherwise.

Good additions:

I particularly like the emphasis on changing to clean gloves and on washing your gloves or hands between steps, as alcohol and chlorine disinfectants are very good for Ebola.

The emphasis on practicing doffing PPE is critical. This has been the step killing HCWs in previous epidemics, like SARS, as it is so easy to contaminate yourself. Most recently, this has been emphasized by Médecins Sans Frontières (MSF, or Doctors Without Borders), cited in the Annals of Internal Medicine.

I would remind readers of two other points: First, Overprotection Does Not Equal Protection, as noted by Dr. Eli Perencevich. And “Requiring HazMat suits and respirators will probably decrease the frequency of provider–patient contacts, inhibit providers' ability to examine patients, and curtail the use of diagnostic tests...Using extra gear inflates patients' and caregivers' anxiety levels, increases costs, and wastes valuable resources."

Secondly, remember that we don’t need high tech to control Ebola. This has been demonstrated repeatedly. In the 1995 Kikwit oubreak, we learned that transmission was not airborne and that the epidemic could be controlled with very basic personal protective equipment. PAPRs and HazMat suits may, in fact, be more of a risk, given they complicate care.This WHO manual from 1995 is a great visual explainer on protecting yourself from hemorrhagic fevers when you have limited resources.


I hope this post will help quell some of the growing and unnecessary anxiety regarding Ebola and that we can soon return our focus to the bigger threat—superbugs like the Klebsiella that even kills patients at premiere hospitals.

These Ebola cases should be regarded as a teaching moment and a time to call for necessary changes and a reinvigoration of infection control and epidemiology practices. If the CDC and hospital administrators fail to capitalize upon the impetus for better infection control that we are now seeing with Ebola, that will be a far greater tragedy.

I again call on the CDC and State Health Departments to focus some of their attention away from bioterrorism preparedness and Ebola, to remedial infection control 101. If all health care workers are trained—and retrained, demonstrating proficiency in practice—we can help protect the country from any infectious diseases threat.

*I have been unable to reach CDC for clarification

Resource for PPE:

Emory University on doffing PPE


Graphics: all from CDC except proposed Risk-based Ebola isolation algorithm, courtesy Scientific American editor Jen Christiansen.

Special thanks to Larry Lynam (@scopedbylarry) for kindly reviewing and critiquing the proposed algorithm.

“Molecules to Medicine” banner © Michele Banks