Vous êtes sur la page 1sur 6

ORIGINAL CONTRIBUTION

Transmission of Mycobacterium tuberculosis


From Medical Waste
Kammy R. Johnson, DVM, PhD Context Washington State has a relatively low incidence rate of tuberculosis (TB)
Christopher R. Braden, MD infection. However, from May to September 1997, 3 cases of pulmonary TB were re-
ported among medical waste treatment workers at 1 facility in Washington. There is
K. Lisa Cairns, MD, MPH no previous documentation of Mycobacterium tuberculosis transmission as a result
Kimberly W. Field, RN, MSN of processing medical waste.
A. Craig Colombel, BS Objective To identify the source(s) of these 3 TB infections.
Zhenhua Yang, PhD Design, Setting, and Participants Interviews of the 3 infected patient-workers and
their contacts, review of patient-worker medical records and the state TB registry, and
Charles L. Woodley, PhD collection of all multidrug-resistant TB (MDR-TB) isolates identified after January 1, 1995,
Glenn P. Morlock, MS from the facilitys catchment area; DNA fingerprinting of all isolates; polymerase chain
reaction and automated DNA sequencing to determine genetic mutations associated with
Angela M. Weber, MS drug resistance; and occupational safety and environmental evaluations of the facility.
A. Yvonne Boudreau, MD, MSPH Main Outcome Measures Previous exposures of patient-workers to TB; verifica-
Thomas A. Bell, MD, MPH tion of patient-worker tuberculin skin test histories; identification of other cases of TB
in the community and at the facility; drug susceptibility of patient-worker isolates; and
Ida M. Onorato, MD, MPH potential for worker exposure to live M tuberculosis cultures.
Sarah E. Valway, DMD, MPH Results All 3 patient-workers were younger than 55 years, were born in the United
Paul A. Stehr-Green, MPH, DrPH States, and reported no known exposures to TB. We did not identify other TB cases.
The 3 patient-workers isolates had different DNA fingerprints. One of 10 MDR-TB catch-

M
YCOBACTERIUM TUBERCU- ment-area isolates matched an MDR-TB patient-worker isolate by DNA fingerprint pat-
losis is usually transmit- tern. DNA sequencing demonstrated the same rare mutation in these isolates. There was
ted when persons with no evidence of personal contact between these 2 individuals. The laboratory that ini-
pulmonary tuberculosis tially processed the matching isolate sent contaminated waste to the treatment facility.
(TB) aerosolize bacteria by coughing, The facility accepted contaminated medical waste where it was shredded, blown, com-
pacted, and finally deactivated. Equipment failures, insufficient employee training, and
sneezing, speaking, or singing.1 Sus-
respiratory protective equipment inadequacies were identified at the facility.
ceptible persons inhale the aerosol-
ized bacteria, which then can implant Conclusion Processing contaminated medical waste resulted in transmission of
deep within the lung and establish in- M tuberculosis to at least 1 medical waste treatment facility worker.
JAMA. 2000;284:1683-1688 www.jama.com
fection. However, novel methods of
transmission have been reported, for ex-
ample, nosocomial infections result- persons,9 which was less than the na- at least isoniazid and rifampin) is un-
ing from the use of poorly cleaned, con- tional TB incidence rate of 7.4 cases per common in Washington State. Five or
taminated bronchoscopes2,3; infections 100 000 persons.10 Furthermore, mul- fewer cases have been reported annu-
due to the aerosolization of bacteria tidrug-resistant TB (isolates resistant to ally during 1992-1997.9 This report
during vigorous wound4 irrigation; dis-
Author Affiliations: Epidemic Intelligence Service, of Surveillance Hazard Evaluations and Field Studies,
posal of peritoneal dialysate5; and au- Division of Applied Public Health Training, Epidemiol- National Institute for Occupational Safety and
topsy6,7 and embalming procedures.8 To ogy Program Office (Drs Johnson and Cairns), Divi- Health, Atlanta, Ga and Denver, Colo (Ms Weber
our knowledge, transmission of M tu- sion of Tuberculosis Elimination, National Center for and Dr Boudreau); and Lewis County Health Officer,
HIV, STD, and TB Prevention (Drs Braden, Onorato, Chehalis, Wash (Dr Bell).
berculosis as a result of processing medi- and Valway), and Division of AIDS, STD, and TB Dr Johnson is now with the Health Studies Branch,
cal waste has not been documented. Laboratory Research, National Center for Infectious Division of Environmental Hazards and Health Evalu-
Diseases (Dr Woodley and Mr Morlock), Centers for ation, National Center for Environmental Health, Cen-
In 1997, Washington State had a TB Disease Control and Prevention, Atlanta, Ga; Wash- ters for Disease Control and Prevention.
incidence rate of 5.4 cases per 100 000 ington State Department of Health, Olympia and Deceased.
Seattle (Drs Johnson, Stehr-Green, and Cairns, Ms Corresponding Author and Reprints: Kammy
Field, and Mr Colombel); Central Arkansas Veterans Johnson, DVM, PhD, Centers for Disease Control and
For editorial comment see p 1701. Health Care System, Little Rock (Dr Yang); Hazard Prevention, 1600 Clifton Rd NE, MS E-23, Atlanta, GA
Evaluation and Technical Assistance Branch, Division 30333 (e-mail: kdj8@cdc.gov).

2000 American Medical Association. All rights reserved. (Reprinted) JAMA, October 4, 2000Vol 284, No. 13 1683

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015


TRANSMISSION OF M TUBERCULOSIS FROM MEDICAL WASTE

summarizes an investigation of 3 TB Control and Prevention, National Tu- viewed the facilitys work and safety
cases reported from May through Sep- berculosis Genotyping and Surveil- policies and procedures, met with lo-
tember 1997. All 3 patients lived in a lance Network regional laboratory da- cal health care providers, reviewed em-
county (population approximately tabase of M tuberculosis isolates. ployee medical records, and photo-
67 000) in Washington State that had a To identify M tuberculosis drug- graphed and videotaped the plants
history of 3 or fewer cases of TB per year resistant mutations, regions of the RNA methods and procedures.14,15 In addi-
for the past decade,9 and the patients had polymerase b subunit (rpob) gene and tion, in their evaluation, the NIOSH
been employed as workers in a medical the entire pyrazinamidase (pncA) gene team used a variety of environmental
waste treatment facility. Our investiga- were amplified using polymerase chain sampling methods to assess the poten-
tion sought to identify the source of the reaction. The resulting amplicons were tial for aerosolization of medical waste
infections and to determine if expo- sequenced using an ABI 373XL auto- during processing.14
sure to medical waste resulted in the mated DNA sequencer (Applied Biosys-
transmission of M tuberculosis. tems, Inc, Foster City, Calif). The pan- Investigation of Clinical Laboratory
susceptible strain M tuberculosis H37Rv Waste Disposal Methods
METHODS was used as a control. The sequences of To determine if worker exposure to live
Investigation of Patient-Workers the patient-worker isolates and the con- M tuberculosis cultures was possible, we
We interviewed each patient-worker trol strain were compared with the pub- conducted a telephone survey of all
and reviewed their medical records to lished sequences of these genes using the clinical laboratories in Washington State
determine their previous exposures to Sequence Navigator software (version approved to perform acid-fast bacilli
TB and to verify their histories of tu- 1.0.1; Applied Biosystems, Inc). (AFB) testing. Personnel at each labo-
berculin skin tests. We reviewed the ratory were questioned about types of
state TB registry for 1996 and 1997 to Investigation diagnostic testing performed, current
identify other cases of TB in the com- of Patient-Worker Contacts and past laboratory waste decontami-
munity. We also matched the current We interviewed the household, social, nation procedures, and waste disposal
and former (1996-1997) employee ros- and coworker contacts of the patient- methods.16
ters of the medical waste treatment fa- workers and reviewed their medical re-
cility to the state TB registry to iden- cords. All contacts underwent tuber- RESULTS
tify other employees diagnosed with TB. culin skin test screening with 5 units Patient-Worker Investigation
of purified protein derivative.13 Per- All 3 patient-workers were white, US
Laboratory Investigation sons who had recent contact with a pa- born, between the ages of 28 and 52
of Patient-Worker Isolates tient-worker underwent a second skin years old, and seronegative for the hu-
We submitted M tuberculosis isolates test 3 months after the last exposure. man immunodeficiency virus (HIV). All
from each of the patient-workers for We defined a positive skin test result 3 worked at the medical waste treat-
susceptibility testing. Laboratory re- as 5 mm or larger induration and a tu- ment facility, for various amounts of
cords and specimen handling meth- berculin skin test conversion as an in- time between 1992 and 1997. None of
ods were reviewed for all laboratories crease in induration of 5 mm or more them had known exposures to any
involved in processing the isolates to compared with a negative test result other persons with active TB, and none
identify potential laboratory cross- within the previous 2 years.13 of the 3 had documented results of prior
contamination or mislabeling of speci- tuberculin skin tests. The first patient-
mens.11 Representative isolates from Environmental Investigations worker identified (patient-worker 1)
each patient-worker underwent DNA of the Medical Waste worked at the medical waste treat-
fingerprinting using IS6110-based re- Treatment Facility ment facility for 4.5 years. Patient-
striction fragment length polymor- The Washington State Department of worker 1 reported a productive cough
phism, according to published stan- Labor and Industries Division of In- in December 1996 and was evaluated
dards.12 M tuberculosis isolates resistant dustrial Safety and Health, Olympia, for pneumonia in April 1997. Infil-
to isoniazid, rifampin, and streptomy- Wash, performed a safety and health trates and cavitary lesions were seen on
cin collected from patients after Janu- evaluation of the facility and, subse- chest radiograph, smears of sputum
ary 1, 1995, from the catchment re- quently, the National Institute for Oc- samples contained numerous AFB,
gion of the medical waste facility cupational Safety and Health (NIOSH) and sputum cultures grew isoniazid-
(medical waste was received from Or- evaluated the potential for occupa- resistant M tuberculosis. Patient-
egon, Washington, Idaho, and British tional exposure to M tuberculosis from worker 1 did not report other risk
Columbia) also underwent DNA fin- processing the medical waste. The in- factors for TB infection.
gerprinting. The DNA fingerprints from vestigators conducted walk-through Patient-worker 2 worked at the facil-
the patient-worker isolates were cross- surveys of the facility, interviewed em- ity for 6 months. Patient-worker 2
matched with the Centers for Disease ployees, observed work practices, re- described the onset of cough and fever
1684 JAMA, October 4, 2000Vol 284, No. 13 (Reprinted) 2000 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015


TRANSMISSION OF M TUBERCULOSIS FROM MEDICAL WASTE

during November and December 1996.


Table 1. Laboratory Characteristics of Mycobacterium tuberculosis Isolates From Medical
After hearing of patient-worker 1s di- Waste Workers
agnosis, patient-worker 2 sought a tu- Patient-Worker Isolate Antimicrobial DNA Fingerprint
berculin skin test in June 1997 and the No. Resistance Pattern Characteristics
results were positive (.15 mm). Chest 1 Isoniazid 19 bands*
radiographs revealed infiltrates, smears 2 None 15 bands
of sputum samples were positive for 3 Isoniazed, rifampin, streptomycin 11 bands
AFB, and sputum cultures grew M tu- *Bands refer to the number of IS 6110 hybridizing fragments in the restriction fragment length polymorphism pattern.
berculosis that was sensitive to all drugs
tested. Patient-worker 2 had a history of
incarcerations in a county jail and a psy- were identified from the National Tuber-
Figure. DNA Fingerprinting Analyses
chiatric hospitalization but had not had culosis Genotyping and Surveillance of Mycobacterium tuberculosis Isolates
previous tuberculin skin tests. Network regional database.
Patient-worker 3 worked at the medi- We identified 10 M tuberculosis iso- S 1 2 3 4 S
cal waste facility for 2.5 years after report- lates from the medical waste process-
15.0
ing the onset of a productive cough at ing facilitys catchment region with
the end of July 1997. Patient-worker 3 resistance patterns that matched the iso-
was screened as an employee contact of lates from patient-worker 3 (ie, resis- 7.5
patient-workers 1 and 2 and in August tant to isoniazid, rifampin, and strep-
1997, had a positive tuberculin skin test tomycin. One regional isolate had a
5.3
result (17 mm). The chest radiograph DNA fingerprint pattern that was iden-
showed bilateral apical densities. Smears tical to that from patient-worker 3
of sputum samples did not reveal AFB, (Figure). Approximately 96% of ri-
4.0
but sputum cultures grew M tuberculo- fampin-resistant M tuberculosis iso-
sisthatwasresistanttoisoniazid,rifampin, lates possess a mutation within an 81
and streptomycin. Patient-worker 3 did base pair region of the rpob gene.17 The
3.1
not recall exposure to patients with ac- isolate from patient-worker 3 and the
tive TB other than patient-workers 1 and matching regional isolate both had the
2. He had a history of substance abuse same mutation in codon 516 (GAC
2.5
andunderwentinpatientsubstanceabuse GTC) resulting in the substitution of
treatmentseveralyearsbeforebutwasnot valine for aspartic acid. This particu-
Kilobase

screened for TB at that time. lar mutation is found in approxi-


2.0
All 3 patient-workers were resi- mately 6% of rifampin-resistant iso-
dents of the community for more than lates. 17 The entire pncA gene was
7 years. The only cases of TB reported sequenced, and both isolates had the
from the county for 1997 were these 3 same silent mutation in codon 38 (GCG 1.6
patients. None of the former employ- GCC). Silent mutations do not
ees of the facility had been reported as change the amino acid sequence of the 1.4

having a case of TB. translated protein and are rarely seen


in M tuberculosis.18 1.2
Laboratory Investigation
After review of laboratory records and Patient-Worker Contact
procedures of the 5 laboratories involved Investigation 1.0

in processing the M tuberculosis isolates We performed tuberculin skin tests for


from the 3 patient-workers, no evi- all patient-worker contacts (TABLE 2).
dence of laboratory cross-contamina- No household contacts had positive tu- 0.8

tion or mislabeling of isolates was found. berculin skin test results. One social
The laboratory characteristics of iso- contact had a positive skin test result;
lates from each patient-worker are sum- this contact had not had previous tu-
marized in TABLE 1. The isolates from the berculin skin tests but had a history of
3 patient-workers showed different anti- substance abuse. A follow-up chest ra- S indicates molecular weight standard. DNA finger-
microbial resistance patterns (Table 1), diograph showed no abnormalities. prints of isolates from patient-worker 1 (lane 1), pa-
tient-worker 2 (lane 2), patient-worker 3 who had mul-
and their DNA fingerprints were dis- All 29 employees (7 clerical and 22 tidrug- resistant tuberculosis (lane 3), and patient with
tinct from one another (FIGURE). No waste workers) of the medical treat- multidrug-resistant tuberculosis from the facility catch-
ment region (lane 4).
isolates with matching fingerprints ment facility were screened as co-
2000 American Medical Association. All rights reserved. (Reprinted) JAMA, October 4, 2000Vol 284, No. 13 1685

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015


TRANSMISSION OF M TUBERCULOSIS FROM MEDICAL WASTE

pervious shoes and gloves, and a sup-


Table 2. Investigation of the Contacts of Medical Waste Workers Infected With
Tuberculosis* plied-air hooded respirator. All patient-
No./Total (%) workers worked with the waste prior to
the electrothermal deactivation decon-
Coworker Contacts tamination step.
Clerical Waste Household and Close Total No. of Contacts
Patient-Worker Workers Workers Social Contacts per Patient-Worker Environmental Investigation of the
1 2/7 (29) 11/22 (50) 0/10 (0) 13/39 (33) Medical Waste Treatment Facility
2 ... ... 0/7 (0) 0/7 (0) A safety flap within the in-feed chute,
3 0/5 (0) 0/9 (0) 1/17 (6) 1/31 (3) which was designed to prevent waste
Total 2/7 (29) 11/22 (50) 1/34 (3) 14/77 (18) from being thrown back onto the plant
*No. of contacts with tuberculin skin test results greater than 5 mm divided by the No. of contacts tested. Ellipses
indicate no contacts. floor, had been missing for 2 to 5
Coworker contacts were included in patient-worker 1 contact investigation. years.14,15 Employees reported that when
Coworker contacts for patient-worker 3 were also tested as contacts for patient-workers 1 and 2.
the shredding equipment became
clogged, waste particles reversed direc-
worker contacts of patient-workers 1 Containers are manually emptied into tion and vented from the in-feed chute
and 2 (Table 2). Ten workers had a the in-feed chute and placed on an- onto the plant floor. Employees re-
positive skin test result, and an addi- other conveyor to be washed with steam ferred to this occurrence as blow-
tional 3 had skin test conversions; 4 of or hot chlorinated water. The process- back. Employees did not consistently
these 13 workers had previous risk fac- ing equipment shreds the waste and shower following their shift, nor did they
tors for TB. None of the 10 workers with blows it through the system with fans, decontaminate themselves when leav-
positive tuberculin skin test results had filling vessels with approximately 225 kg ing the enclosed containment room.14
had prior skin tests. All contacts with of waste. Shredded waste is compacted Employees did not fully understand
positive skin test results had normal in the vessels by a hand-operated hy- the potential health risks (eg, needle
chest radiographic findings. draulic press. The vessels are then manu- injuries and waste spills) associated with
ally capped and guided to a conveyor exposures to medical waste nor did they
Medical Waste Treatment Facility that delivers them to a radiofrequency report all such occurrences to super-
The medical waste treatment facility be- oven that decontaminates the waste us- visors.14 Employees also reported uncer-
gan operating in January 1992. The fa- ing a heating process referred to as elec- tainty about the correct way to don and
cility received waste from hospitals, clini- trothermal deactivation. Processed ves- remove protective clothing, and they did
cal laboratories, and medical and dental sels exit the containment room and are not always wear it appropriately. The
clinics in Oregon, Washington, Idaho, probed to determine a core tempera- airline respirator system used inside the
and British Columbia. The facility did not ture. Vessels not reaching a core tem- containment room did not meet the
require incoming waste to be decontami- perature of at least 194F (90C) are requirements for NIOSH approval.
nated prior to receipt and processing; the returned to the oven for reheating. Pro- Workers reported they had responded
packaging and interstate shipping of the cessed waste is then hauled from the fa- to spills without respiratory protec-
waste met published guidelines19 and le- cility by truck to a landfill. tion, that they had repackaged con-
gal requirements. The 3 patient-workers performed dif- taminated waste, and that the doors
The treatment facility is a 13 500- ferent tasks, but they worked in close between the contained processing
square-foot building with an 800- proximity to one another in the facil- equipment and the plant floor were left
square-foot, 2-story steel-walled con- ity; all workers shared a common break open during waste processing. The
tainment room located in the center of room. Patient-worker 1 washed con- NIOSH evaluation documented the
the plant floor where the processing tainers after they were emptied. Patient- blowback phenomenon using smoke
equipment is located.14,20 The contain- worker 2 emptied containers into the tubes and video equipment.14
ment room is designed to operate under processing equipment at the in-feed
negative air pressure relative to the plant station and also provided assistance Clinical Laboratory Waste Disposal
floor. Waste workers perform their to patient-worker 3 numerous times All 44 laboratories licensed to process
duties at various locations around the throughout the day. In accordance with AFB specimens in Washington State were
plant floor or in the containment room. the facilitys task-specific personal- surveyed by telephone. Forty-one labo-
A second-floor office space provides a protective equipment policy, respira- ratories (93%) sent all waste to an off-
work area for clerical staff. tory protection was not required for the site medical waste disposal company,
Containers of waste are manually un- positions of patient-workers 1 and 2. Pa- including the facility where the 3 patient-
loaded from delivery trucks onto a con- tient-worker 3 operated the hydraulic workers worked. Three laboratories (7%)
veyor that delivers them to the in-feed press within the sealed containment incinerated all or nearly all waste on site.
station of the processing equipment. room while wearing a Tyvek suit, im- Of the 41 laboratories that contracted for
1686 JAMA, October 4, 2000Vol 284, No. 13 (Reprinted) 2000 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015


TRANSMISSION OF M TUBERCULOSIS FROM MEDICAL WASTE

off-site waste disposal, 21 (51%) per- First, no evidence indicated that trans- tional group previously identified to be
formed some decontamination on site. mission to the 3 medical waste work- high risk for clinically active TB.23 How-
However, 20 laboratories (49%) did not ers occurred as a result of exposure to ever, it is unknown to what extent ex-
decontaminate laboratory waste, includ- cases occurring in the community. DNA posure to infectious medical waste
ing stocks and cultures, prior to ship- fingerprinting data also excluded the played a role in this finding. The 3 per-
ment. Nineteen (95%) of these labora- possibility that transmission occurred sons with documented tuberculin skin
tories shipped waste to processing from one patient-worker to another. test conversions may have been in-
facilities and 1 (5%) to a landfill.16 Second, an unusually high percent- fected by patient-worker 1, who worked
age (45%) of the coworker contacts of for several months while infectious. The
Traceback Investigation the 3 patient-workers had positive tu- coworkers with skin test conversions
The regional isolate that matched the berculin skin test results, especially con- also had brief contact with patient-
isolate of patient-worker 3 was from a sidering there was no evidence that pa- worker 2 while potentially infectious.
patient who was foreign-born, had im- tients transmitted TB to their household Four of the persons with positive skin
migrated to the United States in 1994 contacts. test results acknowledged risk factors
and had known exposures to multi- Third, in Washington State, 49% of prior to the outbreak that increased
drug-resistant TB. This patient under- laboratories routinely shipped viable their potential for exposure to TB.
went testing for symptoms consistent M tuberculosis cultures to medical waste This investigation revealed that in-
with TB in early 1996. At that time, an disposal facilities, including this facil- fectious laboratory waste (eg, stocks and
aspirate specimen collected from a cu- ity, despite Centers for Disease Con- cultures) are frequently discarded into
taneous lower limb lesion was culture- trol and Prevention recommendations the solid medical waste stream in Wash-
positive for M tuberculosis; lung bi- to decontaminate such material prior ington State. Our survey was limited to
opsy and bronchoalveolar wash to disposal.19 Washington State; this finding may not
specimens were AFB smear- and cul- Fourth, the environmental investiga- be true for other states. Medical waste
ture-negative. There was no evidence tions of the facility support the plausi- was identified as the source for patient-
of infectious pulmonary disease. bility of workers being exposed to con- worker 3s infection only because he was
The laboratory that handled the taminated waste in ways that put them infected with an organism with a rare
matching isolate had contracted exclu- at risk for infection. One study reported pattern of drug resistance that limited
sively with the medical waste treat- that bacteria (eg, Bacillus subtilis var the number of possible sources. Be-
ment facility for waste removal since Au- niger), particularly dry spores, added to cause multidrug-resistant TB rarely oc-
gust 1992. Furthermore, the laboratory the waste prior to processing were curred in the treatment facilitys catch-
did not decontaminate cultures of M tu- released from equipment during the com- ment region, cultures had been banked
berculosis with growth on solid media paction process.21 Therefore, all work- since January 1995, permitting com-
prior to discarding them as medical ers in the processing area of the plant may parison of DNA fingerprints from these
waste. A solid media culture with growth have been exposed to aerosols released isolates with the one from patient-
of the matching isolate was discarded in from the shredding process. worker 3. This approach was not pos-
mid-1996 from this laboratory. Speci- Finally, an isolate from patient- sible for the more common isolates from
mens from patient-worker 3 were never worker 3 had a DNA fingerprint pattern the other 2 patient-workers.
processed in this laboratory. that was identical to the pattern for an Documented results of previous tu-
Patient-worker 3 and the patient with isolate from a patient whose only rela- berculin skin tests were not available
the matching isolate were interviewed. tionship to this outbreak was exposure for patient-workers 1 and 2 or for their
Personal contact between the 2 patients to the medical waste stream. That these coworker contacts. These results would
was not identified; they lived more than 2 isolates also share rare mutations have allowed estimation of when their
60 miles from one another and had lived increases the likelihood that both patients exposures occurred. The possibility that
within their respective communities for were infected with the same organism. these patient-workers developed TB
more than 3 years. The patients did not The patient whose isolate matched from previous infections cannot be ex-
have common workplaces, hobbies, so- patient-worker 3s isolate did not have cluded. However, based on the annual
cial groups, or medical care providers. infectious pulmonary TB, and there was incidence of TB in Washington State
no evidence of contact between them. (5.4 cases/100000 persons), we would
COMMENT A substantial percentage (45%) of expect 0.0017 cases of TB annually
To our knowledge, this report is the first the coworker contacts had positive among 32 workers.9 Therefore, the
to document transmission of M tuber- (11-40 mm) tuberculin skin test re- probability of observing 3 cases in 1 year
culosis from occupational exposure to sults. The level of TB infection in this by chance is less than 0.000001.24
infectious medical waste. Five lines of study is 3 to 4 times greater than the The findings of this investigation
evidence from this investigation sup- prevalence for funeral service indus- lead to several recommendations.
port this conclusion. try employees (11.7%),22 an occupa- First, laboratories should review their
2000 American Medical Association. All rights reserved. (Reprinted) JAMA, October 4, 2000Vol 284, No. 13 1687

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015


TRANSMISSION OF M TUBERCULOSIS FROM MEDICAL WASTE

policies and procedures for disposal and cultures as the first step in pro- gens. Medical waste workers who are em-
of viable M tuberculosis stocks and cessing waste. Engineering controls ployed at facilities that accept contami-
cultures. To minimize the risk of that prevent employee exposure to nated waste should be considered for
infection for medical waste workers, potentially infectious waste need to be routine screening for TB.
stocks and cultures should be decon- adopted by medical waste treatment
Previous Presentations: Portions of this study have been
taminated by autoclaving or other facilities. In addition, medical waste presented at the 48th Annual Conference of the Epi-
approved methods prior to disposal, workers need extensive and ongoing demic Intelligence Service, April 19-23,1999, Centers
preferably within the laboratory safety training. Such training should in- for Disease Control and Prevention, Atlanta, Ga;and by
invitation at the Northwest Occupational Health Con-
where they are generated.19 Ideally, cludegeneralinfectioncontrolprinciples, ference, October 1998, Seattle, Wash.
medical waste treatment facilities personal protective equipment use, and Acknowledgment: We thank Donna Osmond, MS, John
Kobayashi, MD, MPH, and Richelle Peterson, RN, Wash-
should not accept contaminated labo- techniquesfor responding to waste spills. ington State Department of Health, Olympia and Seattle;
ratory waste. Facilities that do accept This training will help employees rec- Sheila Bowers, RN, and Susan Bland, RN, Lewis County
Health Department, Chehalis, Wash; Mary Miller, RN,
contaminated waste should consider ognize situations that put them at risk and Margaret Cunningham, RN, MSN, MPH, Wash-
deactivation of contaminated stocks for exposure to potentially viable patho- ington State Department of Labor and Industries.

REFERENCES
1. American Thoracic Society. Diagnostic standards Health, Infectious Disease and Reproductive Health. ington State, 1997 [abstract]. Proceedings of the North
and classification of tuberculosis. Am Rev Respir Dis. 1996;11:23-24. American Regional International Union Against TB and
1990;142:725-735. 10. Reported Tuberculosis in the United States, 1997. Lung Disease; Feb 25-27, 1999; Chicago, Ill .
2. Agerton T, Valway S, Gore B, et al. Transmission Atlanta, Ga: Centers for Disease Control and Preven- 17. Ramaswamy S, Musser JM. Molecular genetic ba-
of a highly drug-resistant strain (strain W1) of My- tion, National Center for HIV, STD, and TB Preven- sis of antimicrobial agent resistance in Mycobacte-
cobacterium tuberculosis: community outbreak and tion, Division of Tuberculosis Elimination; July 1998: rium tuberculosis: 1998 update. Tuber Lung Dis. 1998;
nosocomial transmission via a contaminated broncho- 9-15. 79:3-29.
scope. JAMA. 1997;278:1073-1077. 11. Braden CR, Templeton GL, Stead WW, Bates JH, 18. Kapur V, Whittman TS, Musser JM. Is Mycobac-
3. Michele TM, Cronin WA, Graham NM, et al. Trans- Cave MD, Valway SE. Retrospective detection of labo- terium tuberculosis 15,000 years old? J Infect Dis.
mission of Mycobacterium tuberculosis by a fiberop- ratory cross-contamination of Mycobacterium tuber- 1994;170:1348-1349.
tic bronchoscope: identification by DNA fingerprint- culosis cultures with use of DNA fingerprint analysis. 19. Centers for Disease Control and Prevention, Na-
ing. JAMA. 1997;278:1093-1095. Clin Infect Dis. 1997;24:35-40. tional Institutes of Health. Biosafety in Microbiologi-
4. Hutton MD, Stead WW, Cauthen GM, Bloch AB, 12. van Embden JD, Cave MD, Crawford JT, et al. cal and Biomedical Laboratories. 3rd ed. Washing-
Ewing WM. Nosocomial transmission of tuberculosis Strain identification of Mycobacterium tuberculosis by ton, DC: US Government Printing Office; March 1993.
associated with a draining abscess. J Infect Dis. 1990; DNA fingerprinting: recommendations for a standard- DHHS publication (CDC) 93-8395.
161:286-295. ized methodology. J Clin Microbiol. 1993;31:406- 20. Turnburg WL. Alternative treatment technolo-
5. Matlow AG, Harrison A, Monteath A, Roach P, Balfe 409. gies. Biohazardous Waste: Risk Assessment, Policy,
JW. Nosocomial transmission of tuberculosis (TB) as- 13. Centers for Disease Control and Prevention. Es- and Management. New York, NY: John Wiley & Sons
sociated with the care of an infant with peritoneal TB. sential components of a tuberculosis prevention and Inc; 1996:269-273.
Infect Control Hosp Epidemiol. 2000;21:222-223. control program; and screening for tuberculosis and 21. Emery R, Sprau D, Lao YJ, Pryor W. Release of
6. Templeton GL, Illing LA, Young L, Cave D, Stead tuberculosis infection in high-risk populations: rec- bacterial aerosols during infectious waste compac-
WW, Bates JH. The risk for transmission of Mycobac- ommendations of the Advisory Council for the Elimi- tion: an initial hazard evaluation for healthcare work-
terium tuberculosis at the bedside and during au- nation of Tuberculosis. MMWR Morb Mortal Wkly ers. Am Ind Hyg Assoc J. 1992;53:339-345.
topsy. Ann Intern Med. 1995;122:922-925. Rep. 1995;44(No. RR-11):19-25. 22. Gershon RR, Vlahov D, Escamilla-Cejudo JA, et
7. Ussery XT, Bierman JA, Valway SE, Seitz TA, DiFer- 14. NIOSH Health Hazard Evaluation Report 98- al. Tuberculosis risk in funeral home employees. J Oc-
dinando GT, Ostroff SM. Transmission of multidrug- 0027-2709, October 1998. Centers for Disease Con- cup Environ Med. 1998;40:497-503.
resistant Mycobacterium tuberculosis among per- trol and Prevention, National Institute for Occupa- 23. McKenna MT, Hutton M, Cauthen G, Onorato
sons exposed in a medical examiners office, New York. tional Safetly and Health, Atlanta, Ga. IM. The association between occupation and tuber-
Infect Control Hosp Epidemiol. 1995;16:160-165. 15. Washington State Department of Labor and In- culosis: a population-based survey. Am J Respir Crit
8. Sterling TR, Pope DS, Bishai WR, Harrington S, Ger- dustries, Division of Industrial Safety and Health. Care Med. 1996;154:587-593.
shon RR, Chaisson RE. Brief report: transmission of My- WISHA case file 115232399; November 27, 1997 24. Dean AD, Dean JA, Coulombier D, et al. Epi Info,
cobacterium tuberculosis from a cadaver to an em- (available on request). Version 6: A Word-Processing, Database, and Statis-
balmer. N Engl J Med. 2000;342:246-248. 16. Cairns KL, Johnson KR, Diamond C, et al. Bio- tics Program for Public Health on IBM-Compatible
9. Washington State Tuberculosis Epidemiologic Pro- medical waste disposal practices of laboratories pro- Microcomputers. Atlanta, Ga: Centers for Disease Con-
file. Olympia, Wash: Washington State Department of cessing Mycobacterium tuberculosis specimens, Wash- trol and Prevention; 1995.

1688 JAMA, October 4, 2000Vol 284, No. 13 (Reprinted) 2000 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ by a Fudan University User on 05/13/2015