Preventing Transmission of TB
03/2001
· Infection Control Today - 03/2001: Preventing Transmission of TB
By Judene Bartley, MS, MPH, and Gina Pugliese, RN, MS
Although there has been a steady decline of cases of tuberculosis (TB) in the US, this decrease will continue only if attention is paid to identifying patients with TB, initiating proper treatment, and implementing measures to reduce the risk of transmission to others during periods of infectivity.
Worldwide, there are an estimated eight million new cases of TB each year and three million deaths are attributed to this disease annually.1 In the US, there was a steady decline in the number of new cases of TB from 90 cases per 100,000 population in 1950 to 9.4 cases per 100,000 in 1984. As the number of cases decreased, the US government decreased funding for TB and shifted the money to other public health problems. As a result, many of the TB control efforts were greatly reduced. After 1984, the US had a steady increase in the number of cases of TB. From 1985 to 1992, the number of reported cases of TB increased 20%, resulting in 52,000 excess cases of TB. Persons from 25 to 44 years of age accounted for more than 80% of the total increase in cases during this time interval. Several factors are thought to have contributed to this increase, including the HIV epidemic, large outbreaks of multidrug-resistant TB (MDR-TB), an increase in cases occurring in persons who immigrated to the US from areas of the world with a high prevalence of TB, and an increase in active transmission of TB caused by inadequate healthcare resources. The decline in the number of new cases of TB in the US started in 1993 and has continued through today. The rate of new cases of TB has decreased to 6.4 per 100,000 in 1999, the lowest since US national TB surveillance began in 1953.
In general, persons who become infected with M. tb have a 10% risk for developing active TB during their lifetime. This risk is greatest within the first two years after infection. HIV is the strongest risk factor for progression of latent TB infection to active TB. Persons with latent TB infection who become co-infected with HIV have a 8 to 10% risk per year for developing active TB. HIV-infected persons who are already severely immunosuppressed and who become newly infected with M. tb have an even greater risk for developing active TB.
The probability that a person who is exposed to TB will become infected depends on the concentration of droplet nuclei in the air and the duration of exposure. Characteristics of the TB patient that enhance transmission include:
· Disease of the lungs, airways, or larynx.
· Presence of acid-fast-bacilli in the sputum.
· Failure of patients to cover their mouth or nose when coughing or sneezing.
· Presence of cavitation on the radiograph.
· Inappropriate or short duration of treatment.
· Procedures that induce coughing or aerosolization of M. tb (endotracheal intubation, suctioning, bronchoscopy, sputum induction, suctioning, surgical drainage and irrigation of a TB abscess, and surgical debridement of a tuberculous skin ulcer, administration of aerosolized pentamidine, and autopsy).
Environmental factors that increase risk of transmission include:
· Exposure in a small enclosed space.
· Inadequate local or general ventilation that results in insufficient dilution and/or removal of infectious droplet nuclei.
· Recirculation of air containing infectious droplet nuclei.
The characteristics of the persons exposed to TB that may affect risk of becoming infected are not well defined. In general, persons who have been infected previously with M. tb may be less susceptible to subsequent infection. However, reinfection can occur among those previously infected, especially if they are severely immunocompromised. Vaccination with Bacille of Calmette and Guerin (BCG) probably does not affect the risk of infection; rather it decreases the risk for progressing from latent TB.
Although children who have TB may be less likely than adults to be infectious, they should be evaluated for potential infectiousness with the same criteria as for adults. Pediatric patients that may be infectious include those with laryngeal or extensive pulmonary involvement, pronounced cough, positive sputum for AFB, cavitary TB, or for those whom cough-inducing procedures are performed. The source case for pediatric TB patients often occurs in a member of the child's family; therefore, parents and other visitors of all pediatric TB patients should be evaluated for TB.
Drug Resistance
In the US, from 1993 through 1996, overall resistance to at least isoniazid was 8.4%; rifampin 3.0%; both isoniazid and rifampin (classified as MDR-TB) 2.2%; pyrazinamide 3.0%; streptomycin 6.2%; and ethambutol hydrochloride 2.2%.3 Rates of resistance were significantly higher for case patients with a prior TB episode. Compared with previous US surveys in 1991 and 1992, isoniazid resistance has remained relatively stable. In addition, the percentage of MDR-TB has decreased, although the national trend was significantly influenced by the marked decrease in New York City.
Risk of Nosocomial Transmission
Nosocomial TB has been of significant concern to healthcare workers and the public and few other problems have had such significant impact on hospital epidemiology.4 The US has seen dramatic outbreaks of both multidrug-resistant TB (MDR-TB) and drug-susceptible strains of TB in hospitals with transmission to both patients and health workers.5,6 CDC tracked an outbreak from a specific resistant strain as it spread across the US.7
Nosocomial transmission of TB is not a new issue and it has been known for decades that the risk to healthcare workers is two to ten times greater than that of the general public. However, the magnitude of these recent outbreaks that have involved both patients and healthcare workers has caused significant concern, prompting special infection control measures.
A review of the outbreaks of MDR-TB in the US shows that these outbreaks involved large numbers of cases with a high prevalence of HIV infection. The mortality rate was extremely high and the median interval from TB diagnosis to death was extremely short, the majority being less than four weeks. The high mortality rate in these outbreaks is explained by the severe degree of immunosuppression in many of the patients combined with ineffective treatment for unrecognized drug-resistant disease. Nearly all patients in these outbreaks had M.tb isolates resistant to both isoniazid and rifampin, the two most effective drugs available. In four hospitals and the prison system, the outbreak strain was resistant to seven anti-TB drugs (including streptomycin, ethionamide, cycloserine, kanamycin, rifabutin, and pyrazinamide). At least 20 healthcare workers in these facilities developed active TB, and at least nine workers died.
Some of the major factors contributing to the recent outbreaks of both MDR-TB and drug-susceptible TB in hospitals were breaks in some basic TB control strategies, such as: delays in diagnosis of TB, delays in identification of drug resistance, and delays in initiation of appropriate therapy--all of which resulted in delays in proper isolation and prolonged patient infectiousness.8 Even if a patient was diagnosed with TB, respiratory isolation was often inadequate. For example, isolation rooms were found to have positive rather than negative pressure, air was being recirculated from isolation rooms to other high risk areas, doors to isolation rooms were left open, isolation precautions were discontinued too soon, and healthcare workers did not wear adequate respiratory protection. When appropriate TB control measures were implemented, transmission was significantly reduced or ceased entirely. Unfortunately, many of the interventions were implemented simultaneously, so the effectiveness of specific interventions could not be determined.
Outbreaks in hospitals and prisons illustrate the rapid spread and extent of TB that can occur when people who have undiagnosed or inadequately treated TB, caused by drug-resistant organisms, are brought together with highly vulnerable, immunosuppressed patients, in a densely populated environment in the absence of infection control measures.
There has also been transmission of TB reported in the pediatric setting, related to frequent suctioning and endotracheal intubation, nursing homes for the elderly, and the dental setting, dentist to patients.
In 1994, the US Centers for Disease Control and Prevention (CDC) published a 132-page Guideline for Preventing the Transmission of Mycobacterium TB in Health Care Facilities.6 Follow-up studies by the CDC at several of the hospitals where outbreaks occurred have shown that patient-to-patient and patient-to-healthcare worker transmission was stopped after implementation of these recommended guidelines.8
These nosocomial outbreaks and the concern for healthcare worker safety in the US was the impetus for the US Occupational Safety and Health Administration (OSHA) to get involved and inspect hospitals for compliance with CDC measures to reduce occupational exposure to TB. Non-compliance with the basic requirements can result in significant monetary fines.
Drug susceptibility patterns of M.tb isolates from TB patients treated in the facility should be reviewed to identify frequency and patterns of drug resistance. PPD skin test conversion rates should be analyzed for each department or occupational group and be compared to rates for workers in areas where exposure to TB is unlikely.
Diagnostic Evaluation and Treatment
Prompt and accurate laboratory results are important for the proper treatment of patients with TB. Laboratories must be proficient at processing specimen. Results of acid-fast bacilli (AFB) sputum smears should be available within 24 hours. TB may be more difficult to diagnose among patients with HIV infection because of the nonclassical clinical or radiographic presentation, an impaired response to PPD skin tests, the lower sensitivity of sputum smears for detecting AFB, and the overgrowth of cultures with Mycobacterium avium complex in specimens from patients infected with both M. avium and M.tb.
It will also be important to start empiric therapy as soon as TB is suspected with an appropriate regimen based on the local drug-resistance surveillance data.6,9 The current US recommendation is to begin empiric therapy with four drugs (isoniazid, rifampin, pyrazinamide, and ethambutol, or streptomycin) except in areas where surveillance reveals that the prevalence of primary resistance to isoniazid is less than or equal to 4%.10 In those locations with rates less than or equal to 4%, initial regimens of three drugs (isoniazid, rifampin, and pyrazinamide) are recommended for initial therapy. Treatment guidelines for TB patients with HIV infection and on specific treatment have been updated in light of changing drug resistance trends.11, 12
Prompt identification of patients with TB is essential to TB control efforts.6, 13,14 Unless suspect or confirmed TB patients are identified, it will not matter what other infection control measures are in place. This will require careful evaluation of patients upon their initial encounter with the healthcare system with prompt isolation as soon as TB is suspected on clinical grounds alone and until laboratory and clinical evidence eliminates the diagnosis. The specific procedure for "early identification and isolation of patients" will be based on the prevalence and demographic profile of TB patients in the community. In areas of high prevalence of HIV and TB and because of the difficulty in recognizing TB in patients with HIV, some facilities have found it necessary to isolate all patients with HIV infection that present with clinical symptoms suggestive of TB (e.g., fever, cough) and /or an abnormal chest radiographs until a diagnosis of TB can be ruled out.15
The procedure for "early identification of TB patients" and the definition of "suspect" case will determine the number of isolation rooms that will be needed. There will often be patients who are placed in isolation because of suspect TB and are later found not to have TB. The best strategy is to isolate a suspect patient until TB can be ruled out to prevent possible nosocomial transmission. A number of strategies have been used to increase the availability of rooms that can be used for isolation, such as the use of window exhaust fans/units to create negative pressure or portable or wall-mounted HEPA filtration units.
Giving authority to nursing and physician staff to made independent decisions to isolate patients with suspect TB and policies for automatic isolation of certain patients (e.g., with TB in differential diagnosis) can often reduce delays in initiation of isolation.16 Hospitals that have monitored for compliance with TB control measures, particularly appropriate isolation, have demonstrated reductions in the number of days that potentially infectious patients were not in appropriate isolation.15, 17
The standard method of identifying persons infected with Mycobacterium tb is the Mantoux tuberculin skin test given intradermally with 0.1 ml of 5 tuberculin units of purified protein derivative (PPD) tuberculin.6, 16 It is clear that PPD skin testing of healthcare workers permits early recognition of potential episodes of nosocomial transmission and the opportunity to offer isoniazid or other chemoprophylaxis to workers with skin test conversion. In addition, it is important to identify workers with active TB that pose a risk of infection to patients and others. It may be prudent to do baseline PPD testing of HCWs, particularly in high-risk facilities. The frequency of additional PPD testing will depend on the level of risk in a particular facility.
The prevalence of TB in the facility should be considered when choosing the appropriate cut-point for defining a positive PPD reaction. For example, in facilities with minimal or low risk of TB exposure, an induration of 15 mm may be an appropriate cut-point for workers who have no other risk factors. In other facilities where the risk of TB exposure may be higher, the appropriate cut point may be 10 mm.
All results on PPD testing should be recorded in the worker's health record as well as an aggregate database of all the healthcare worker PPD skin test results. PPD conversion rates should be calculated for the facility as a whole, and if appropriate, for specific areas of the facility and occupational groups. PPD conversion rates should be calculated based on the total number of previously PPD negative HCWs tested in each area or group (i.e., the denominator) and the number of PPD test conversions among HCWs in each area or group (i.e., the numerator).
The potential for variability of skin test conversions with different commercial preparations of PPD is another important issue when evaluating your skin test conversion rates. A number of recent studies have demonstrated a difference in reactivity between various commercial products.18, 19