Critical Care Plus: Study: Respiratory Isolation Measures Underused

Despite measures taken following the resurgence of TB cases in the late 1980s and early 1990s, many health care workers still poorly understand respiratory isolation procedures, says Kevin P. Fennelly, MD, MPH, researcher at the Center for Emerging Pathogens of the New Jersey Medical School in Newark.

Because that resurgence was successfully quashed, today’s health care workers, including intensive care personnel, are less likely to have such expertise because they have less clinical experience with the disease, adds Fennelly. The decrease in TB incidence to historically low levels also creates problems for public health officials working to sustain existing disease control programs and systems because low incidence doesn’t indicate the extent of the efforts required for TB control.

He notes that part of the impetus for the Centers for Disease Control and Prevention (CDC) in-progress revision of its 1994 "Guidelines for Preventing the Transmission of Mycobacterium Tuberculosis in Health Care Facilities"1 is that a good deal of research has been done since the original document.

"More importantly, there’s been a realization that the document was written at a time when there appeared to be some uncertainties in the risk of transmission," he says. Fennelly’s facility is one of three National Tuberculosis Model Centers funded by the CDC that offers training curricula, course materials and technical assistance.

Three Things to Watch

Basically, Fennelly says, controlling M tuberculosis transmission in the ICU boils down to three categories: administrative, engineering and personal respiratory protection.

Administrative controls entail having a system in place so that health care workers are well educated about TB and have a high index of suspicion, Fennelly says. The most important of these is a having a set of policies and procedures for assuring rapid identification and treatment of potential TB-carrying patients.

Patients suspected of having TB should be placed in respiratory isolation in an airborne infection isolation room without delay, Fennelly says, and sputum specimens and other respiratory secretions sent for smears immediately. He notes that there has been a huge problem because results of such diagnostics can come back months after the patient has died from HIV and TB, leaving ICU personnel ignorant of their risks in the interim.

According to recommendations prepared by John A. Jereb, MD, for the CDC’s Division of Tuberculosis Elimination, most low-incidence state TB programs find it hard to assure reports if private medical providers and hospitals send specimens to local hospital laboratories or to out-of-state contract laboratories for testing. Jereb points out that state TB labs require substantial fixed facility and personnel investments that do not decrease even when the TB burden becomes very low.

This situation is similar in the remainder of the country. It puts the TB program at a disadvantage because these laboratories might fail to report critical results promptly to the health department. They also might discard M tuberculosis isolates before subsequent testing, such as DNA fingerprinting, can be done.

For instance, 21 TB cases reported in a small Maine community in 1989-1992 were traced back to a source case diagnosed after an eight-month delay. Fennelly describes a case reported by Anthony Tatandaro, MD, in which an ICU patient whose initial sputum spear was negative infected more than 75% of ICU personnel in attendance after the patient was intubated, mechanically ventilated and had a bronchoscopy procedure. That particular case was notable because the ICU was poorly ventilated, Fennelly notes, adding that the best engineering protection against TB is circulating bad air out and fresh air in. "Some hospitals have used high-efficiency particulate air filters, but they are very expensive and unproven," he says.

Old data that say that ultraviolet germicidal irradiation (UVGI) decreases the amount of viable organisms in the air, and that is supported by yet-to-be-published research in which Fennelly collaborated. It suggests that a top-of-the-line UVGI system works well. However, the best engineering control still is pollution ventilation with at least six and preferably 12 air changes per hour, he says. "We have a lot of theoretical modeling and laboratory-based work, but actual field work is impossible. You can’t run an experiment in which Hospital A has protection against TB but Hospital B doesn’t."

Personal Respiratory Protection

Personal respiratory protection is the most controversial measure, Fennelly says. Particulate respirators are controversial—many people think they are overkill. "There’s a certain tension between traditional infection control practitioners and some of the environmental occupational health people who are relative newcomers to this field," he observes. "I think that’s understandable when you realize that TB is only sporadically infectious."

Fennelly’s main research interest is determining infectiousness of patients. "Many of us have been exposed to TB patients and not become infected," he notes. "That’s probably due as much to the great variability among patients as sources of infection as it is to the variability and susceptibility of individuals who become infected."

Fennelly’s research group, which is funded to study postdeterminants of infectiousness, is currently examining the strength and frequency of cough and viscoelastic properties of secretions to learn which parameters are the most important. Some of the group’s preliminary data suggest that N95 respirators are to be recommended. One objection, he says, is that in most locales N95s cost 75 cents apiece. Though research shows that the most dangerous patients are the ones not yet identified or being treated, Fennelly says that personal respirators may be overkill with patients being treated in isolation rooms with six to 12 air changes an hour.

"I fall down on both sides of the personal respiratory question," Fennelly says. "We know that if you are treating a patient in isolation who’s been on treatment for a while, the decrease in risk from personal protection is very little, but with patients not yet being treated who have a procedure such as a bronchoscopy, the risk is tremendous. With patients undergoing tracheal intubation or cough-producing procedures such as sputum induction or bronchoscopy, the staff should definitely wear some type of respiratory protection." 


1. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health care facilities, 1994. MMWR Morb Mortal Wkly Rep. 1994;43(RR-13):1-132.