19 October 1999
Volume 131
Number 8
Annals of Internal Medicine
Transmission of Tuberculosis in a Jail
Timothy F. Jones, MD; Allen S. Craig, MD; Sarah E. Valway, DMD, MPH; Charles L. Woodley, PhD;
and William Schaffner, MD
Background: Outbreaks of tuberculosis are uncommonly
recognized in jails. In 1996, an increase in active tuberculosis cases was noted among inmates of a large urban jail.
Objectives: To determine the source and extent of a
tuberculosis outbreak in an urban jail and to recommend
control measures.
Design: Retrospective cohort study.
Setting: Urban jail.
Patients: Inmates and guards with tuberculosis.
Intervention: Outbreak evaluation and control.
Measurements: Medical records of inmates and guards
with tuberculosis were reviewed, and inmates were interviewed. DNA fingerprinting was performed on Mycobacterium tuberculosis isolates.
Results: From 1 January 1995 through 31 December 1997,
active tuberculosis was diagnosed in 38 inmates and 5
guards from the jail. Nineteen (79%) of the 24 culturepositive inmates had isolates with DNA fingerprints matching those of other inmates. Isolates from both culturepositive guards matched the predominant inmate strain;
only 6 (14%) of 43 isolates from infected persons in the
community had this pattern. The median length of incarceration of all inmates in the jail was 1 day; the median
length of continuous incarceration before diagnosis of
tuberculosis in inmates was 138 days. Inmates with tuberculosis had been incarcerated a median of 15 times. Fortythree percent of persons in this city with tuberculosis
diagnosed from January 1995 through July 1997 had been
incarcerated in the jail at some time before diagnosis.
Conclusions: Traditional and molecular epidemiologic
investigations suggest that tuberculosis was transmitted
among inmates and guards in an urban jail. Aggressive
measures to screen for active tuberculosis upon incarceration are important for preventing spread of disease in jails
and to the surrounding community.
Ann Intern Med. 1999;131:557-563.
T
uberculosis is a problem in correctional facilities throughout the United States. In 1996,
3.7% of all cases of tuberculosis nationwide occurred among residents of correctional facilities (1).
Although in 1996 the incidence of new cases of
active tuberculosis in the United States was 8.0 per
100 000 persons (1), many prison systems have reported rates of 200 per 100 000 persons and higher
(2–10). Transmission of tuberculosis from prisons
into surrounding communities has been documented
(5), and correctional facilities may be important reservoirs of infection (11–13). One study concluded
that a prison was potentially linked to 9% of a state’s
tuberculosis cases during a 5-year period (5), and
another study indicated that 24% of the tuberculosis
cases in a county were associated with its jail (14).
Prisons house convicts after sentencing, usually
for terms exceeding 1 year (15). In contrast, jails
receive prisoners immediately after arrest and generally house inmates awaiting trial or those sentenced to terms less than 1 year (16). Most jails are
operated by cities or counties and hold inmates
from the local community. In 1993, nearly 10 million inmates were admitted to local jails (16). Only
6% of jails house more than 50% of the nation’s jail
inmates (16). Incarcerated populations have high
rates of substance abuse, HIV infection, tuberculous
infection, low socioeconomic status, and other risk
factors associated with active tuberculosis (15, 17).
At least 14 reports of outbreaks of tuberculosis in
U.S. prisons have been published since 1985 (3, 5–7,
18, 19), but only two published reports have concerned outbreaks in jails (14, 20). Memphis, Tennessee, has the fifth largest jail in the United States
(16). In 1996, the number of reported cases of
tuberculosis diagnosed in inmates from the Memphis jail increased. This report summarizes the results of the ensuing investigation.
See editorial comment on pp 617-618.
For author affiliations and current addresses, see end of text.
© 1999 American College of Physicians–American Society of Internal Medicine
557
Methods
We reviewed the medical records of all persons
in whom tuberculosis was diagnosed from 1 January
1995 through 31 December 1997 while they were
incarcerated in the Memphis Criminal Justice Center (subsequently referred to as the “jail”). We obtained medical records from the jail, the Memphis/
Shelby County Health Department Tuberculosis
Clinic, and the hospital to which inmates were admitted for evaluations. For all inmates with active
tuberculosis, computerized jail records were analyzed to determine dates of incarceration and cell
locations for the 2 years before their diagnosis.
A patient with a confirmed case of active tuberculosis had Mycobacterium tuberculosis isolated from
a clinical specimen or met a clinical case definition
(21). Clinical cases met the following criteria: 1) a
positive result on a tuberculin skin test, 2) signs and
symptoms compatible with tuberculosis (for example, an abnormal, unstable chest radiograph or clinical evidence of current disease), and 3) treatment
with at least two antituberculosis medications. Inmate cases were defined as patients in whom confirmed tuberculosis was diagnosed while they were
incarcerated in the jail or within 3 weeks of transfer
from the jail to another penal facility in the same
city. Persons with culture-positive pulmonary tuberculosis were considered infectious from 6 weeks before collection of the first positive specimen until 2
weeks after initiation of appropriate therapy. A positive tuberculin skin test result was defined as at least
10 mm of induration within 48 to 72 hours after
administration of five tuberculin units of tuberculin
purified-protein derivative by the Mantoux method.
Efforts were made to contact inmates who received a diagnosis of active tuberculosis after 1 January 1995, including those living in the community
or incarcerated in other correctional facilities. A
standardized questionnaire was administered to persons who could be located; the questionnaire asked
Figure 1.
558
about the patient’s lifestyle outside the jail before
diagnosis of tuberculosis (for example, places frequented, behaviors, living and working situations, and
exposure to persons with tuberculosis). Patients were
also asked about incarceration history, possible contacts during incarceration, and activities while in jail.
Jail administrators were interviewed and records
were reviewed to determine baseline information on
the jail population. The average age of inmates was
calculated from a sample of 2552 inmates released
or transferred from the jail on 12 randomly selected
days in 1997. Inmate intake procedures at the jail
were observed by two separate investigators on different days. Records from annual tuberculin skin
test screening of jail staff, performed by the county
health department on site at the jail, were reviewed.
Medical records of staff with confirmed tuberculosis
between 1 January 1995 and 31 December 1997
were also reviewed. The Tennessee Department of
Health registry of all persons in Memphis reported
with active tuberculosis from January 1995 through
July 1997 was cross-matched with the list of persons
incarcerated in the Memphis jail since 1979.
Deoxyribonucleic acid fingerprinting was performed on all available M. tuberculosis isolates from
culture-positive inmates and guards. Fingerprinting
was also done for a sample of community cases,
chosen by selecting every fifth isolate from a list of
all culture-positive cases of tuberculosis in Memphis
from 1 January 1995 through 1 September 1997,
sorted by date of treatment initiation. If no isolate
was available for the case selected, the next community case on the list was substituted. The Tennessee Department of Health Laboratory had processed 99.5% of all specimens from culture-positive
cases reported from Memphis during the study period. Isolates were fingerprinted at the Centers for
Disease Control and Prevention by IS-6110 restriction fragment length polymorphism analysis (22).
Bivariate statistical analyses were performed by us-
Distribution of lengths of incarceration of inmates in the Memphis jail, 1995 to 1997.
19 October 1999
•
Annals of Internal Medicine
•
Volume 131
•
Number 8
results. Inmates who volunteer are counseled about
and tested for HIV at the jail by county health
department representatives. Of the inmates admitted to the jail during this 3-year period, less than
1% were screened for HIV there. Of 1622 HIV
tests performed, 26 (1.6%) had positive results.
ing chi-square tests calculated with Epi Info software (23).
Results
Jail Characteristics
Inmate Tuberculosis Cases
The jail housed approximately 2700 inmates at
any time. More than 173 000 persons were admitted
and discharged during the 3-year period. A mean of
159 inmates were admitted daily; the median length
of stay was 1 day, and 8.3% of inmates stayed more
than 30 days (Figure 1). Of persons admitted to the
jail, 82% had previously been incarcerated there.
The inmate population was 90% black and 90%
male. The mean age of inmates was 32 years.
Inmates were housed on seven floors. Some units
held up to 36 inmates in a single large room; other
units had 18 two-person cells, and the 36 inmates
intermingled for much of the day. Inmates lived and
ate with members of the same unit, although inmates from different units had the potential to intermingle at visitation; during gym, chapel, infirmary, and legal-room (library) visits; and during
transport outside the facility. For security reasons,
inmates were moved frequently within the jail.
Active tuberculosis was diagnosed in 38 jail inmates from 1 January 1995 through 31 December
1997 (Figure 2). Ten cases were diagnosed in 1995,
19 in 1996, and 9 in 1997. The calculated incidence
of active tuberculosis diagnosed in inmates who
were physically in the jail was 274 per 100 000 during this 3-year period.
Inmates with tuberculosis had been in the jail a
median of 15 times before diagnosis. The median
length of continuous incarceration before the diagnosis of tuberculosis was 138 days (range, 10 to 800
days) (Table). By comparison, the median length of
incarceration for all inmates in the jail is 1 day
(mean, 13.4 days). Ten (26%) of the inmates with
active tuberculosis were given the diagnosis during
evaluation for symptoms, 25 (66%) were discovered
as a result of evaluation of a positive tuberculin skin
test result, and 3 (8%) were found during contact
investigations of other cases.
Of the 38 inmates with a diagnosis of tuberculosis, 24 had positive cultures. Sixteen (67%) of the 24
jail inmate isolates of Mycobacterium tuberculosis
Jail Screening Procedures
Routine medical screening at intake to the jail
consisted of two questions: “Are you seeing a doctor for anything?” and “Are you taking any medications?” This screening process took approximately
15 seconds per inmate. If the screening procedure
did not reveal an obvious history of tuberculosis, the
inmate was admitted into the general jail population. Jail protocol called for tuberculin skin test
screening of all inmates still in the facility 10 days
after admission. Inmates with evidence of possible
tuberculosis on initial screening, in subsequent medical visits, or on tuberculin skin test screening were
transported to a hospital emergency department or
the county health department tuberculosis clinic for
chest radiographs and evaluation. Those thought to
have possible active tuberculosis were then transferred to a local hospital for isolation and completion of evaluation. Inmates given a diagnosis of
active tuberculosis were returned to the jail after
three sputum smears were negative for acid-fast
bacteria. The jail does not have radiography facilities or negative-pressure isolation rooms.
From 1995 through 1997, 173 815 inmates were
admitted to the jail; 13 239 (7.8%) inmates underwent tuberculin skin testing. These inmates represented 36% of the target population of inmates
incarcerated for more than 10 days. Of tuberculin
skin tests placed, 10 110 (74%) were read; 431
(4.3%) of these were reported as yielding positive
19 October 1999
Figure 2. Cases of active tuberculosis disease diagnosed in guards
(top) and inmates (bottom) from the Memphis jail, 1 January 1995
through 31 December 1997, by quarter in which disease was diagnosed. White bars indicate persons with isolates with DNA fingerprints
matching the predominant strain in the jail, striped bars indicate those with
a different DNA fingerprint, and black bars indicate culture-negative persons.
•
Annals of Internal Medicine
•
Volume 131
•
Number 8
559
Table.
Characteristics of the 38 Inmates in Whom
Tuberculosis Was Diagnosed in the Memphis
County Jail, 1 January 1995 to 31 December 1997
Characteristic
Value
Median age (range), y
Men, n (%)
Ethnicity, n (%)
Black
Non-Hispanic white
Born in the United States, n (%)
Median duration of continuous incarceration
in the jail before diagnosis (range), d
HIV status, n (%)
Positive
Negative
Unknown
Site of disease, n (%)
Pulmonary
Nonpulmonary
Findings on chest radiography, n (%)
Cavitary
Infiltrate/densities
Adenopathy
Effusion
Normal
Laboratory results, n (%)
Sputum smear–positive
Culture-positive
DNA fingerprint matched predominant jail
strain (24 isolates tested)
Drug-resistant organism
34 (19 –52)
38 (100)
37 (97)
1 (3)
38 (100)
138 (10 – 800)
5 (13)
27 (71)
6 (16)
34 (89)
4 (11)
4 (11)
22 (58)
8 (21)
2 (5)
2 (5)
11 (29)
24 (63)
16 (67)
0
had identical DNA fingerprints, with 11 matching
bands on IS-6110 restriction fragment length polymorphism testing (Figure 3). Among the 8 inmates
with fingerprints that did not match the predominant strain, 3 had isolates with an identical restriction fragment length polymorphism pattern with 12
bands; thus, 19 (79%) of 24 culture-positive inmates
had isolates that matched those of two or more
other inmates.
Restriction fragment length polymorphism testing
was done on 43 isolates from randomly selected
community cases from the same city as the inmate
cases. Of these, 28 (65%) were from men, 30 (70%)
were from black persons, and 19 (44%) were from
black men. Only 6 (14%) of 43 isolates from persons with active tuberculosis in the community had
the 11-band epidemic pattern; 4 of these persons
had been previously incarcerated in the jail. Culture-positive cases among inmates were significantly
more likely to have isolates with a DNA fingerprint
matching the predominant strain in the jail than
were cases among persons in the community (P ⬍
0.001). All M. tuberculosis isolates from culture-positive inmates were susceptible to all drugs tested.
Of the 19 culture-positive inmates with isolates
matching those of other inmates, 3 (16%) were
housed in the same unit with another inmate who
had an identical isolate during their presumed infectious period. Ten (53%) had been housed on the
same floor as another inmate with a matching isolate during their infectious period. All inmates with
560
19 October 1999
•
Annals of Internal Medicine
•
active tuberculosis had been incarcerated in the jail
at the same time as other infected inmates during
their potential infectious period. The periods of incarceration of the 16 inmates with the predominant
M. tuberculosis strain are shown in Figure 4.
Guard Tuberculosis Cases
Active tuberculosis was diagnosed in five jail staff
members during the 3-year period; all were guards.
Two guards were culture-positive, and both had
strains identical to the predominant inmate strain.
Guards are frequently rotated to various duty stations, and all have direct contact with inmates. Both
of the culture-positive guards were employed during
periods overlapping the incarcerations of several inmates with matching isolates (Figure 4). Of the five
guards with tuberculosis, four (80%) were female
and all were black, with a mean age of 34 years and
a mean length of employment of 47 months. All
lived in the same city in which the jail is located. Of
approximately 790 jailors at the facility, 56% were
female, and 90% were black; the mean age was 38
years, and the mean length of employment was 79
Figure 3. Autoradiograph of IS-6110 restriction fragment length
polymorphism patterns of Mycobacterium tuberculosis isolates.
Lanes 1 and 11 are standards. Lanes 4 through 7 and 9 are from inmates
whose isolates had the 11-band pattern matching that of 12 inmates and
two guards. Lanes 2 and 3 are from two of three inmates with an identical
12-band pattern. Lane 8 is from an inmate whose isolate did not match
those of other jail cases, and lane 10 is from a person living in the community.
Volume 131
•
Number 8
months. None of these characteristics differed significantly between the group with tuberculosis and
that without.
The results of annual tuberculin skin test screening of jail staff were available for 2 of the years in
this study period. In October 1996, 706 staff were
evaluated; 128 (18.1%) had previously had a positive test result, and 19 (2.7%) were reported with
newly positive tuberculin skin test results. In October 1997, 686 jailors were evaluated; 537 (78%) had
a tuberculin skin test reading of 0 mm, 140 (20%)
had previously had a positive test result, and 8
(1.2%) had a skin test that had converted to 10 mm
or greater since the preceding year. Those with a
negative tuberculin skin test result had been employed at the jail for an average of 75 months,
compared with 92 months for those with a positive
test result and 94 months for convertors. All staff
with documented conversions in 1997 were guards.
Figure 4. Periods of incarceration, during the 2 years before diagnosis, of the 16 jail inmates (black lines) whose Mycobacterium
tuberculosis isolates had the predominant DNA fingerprint and periods of employment of the two jail guards (striped lines) with the
same strain. Boxes indicate periods in the jail when persons were presumably infectious. Inmate P was not considered infectious.
Discussion
Patient Interviews
Several lines of evidence suggest that M. tuberculosis was transmitted in this jail. Thirty-eight cases
of active tuberculosis were diagnosed in inmates
during the 3-year period. Of the culture-positive
inmate cases, 67% had isolates with a single identical fingerprint pattern on restriction fragment
length polymorphism analysis; only 14% of isolates
from persons with active tuberculosis in the community had this pattern. Of culture-positive inmates,
79% had isolates matching those of two or more
other inmate cases. Inmates with active tuberculosis
had median lengths of incarceration significantly
longer than those of the overall jail population and
had been incarcerated a median of 15 times. No
common exposures among inmates with tuberculosis
outside the jail could be identified. In addition, five
guards were given a diagnosis of active tuberculosis
during this period, and both guards with positive
cultures had isolates identical to the predominant
inmate strain.
Molecular epidemiology played an important role
in recognizing and defining this outbreak. Restriction fragment length polymorphism clustering is reported to be highly predictive for epidemiologic
connections among patients in large, crowded, or
shifting populations and in settings such as prisons
and shelters (24). Additional testing of the predominant inmate strain for polymorphism in the genes
encoding catalase-peroxidase and the A subunit of
gyrase identified the strain as belonging to one of
two genotypic groups found to be associated with
large case clusters (25). The use of DNA fingerprinting to complement traditional epidemiologic
techniques may be particularly useful in jails, where
the barriers to other methods of investigation are
higher than in many other settings.
Of the 38 inmates with active tuberculosis diagnosed since 1 January 1995, 15 (39%) could be
contacted. All 15 agreed to be interviewed. Nine
were inmates with isolates matching the predominant strain, 4 had a different strain, and 2 were
culture negative. No common gathering areas (for
example, residences, workplaces, hotels, social service agencies, parks, clubs, bars, and street corners)
or contacts outside the jail were identified by any
two inmates. However, 14 (93%) of the inmates had
participated in visitation or in chapel, library, or
gym visits while in the jail.
Health Department and Jail Registry
Cross-Matching
Cross-matching of the health department tuberculosis case registry and jail records indicated that,
since 1995, 43% of persons with tuberculosis reported in Memphis had been incarcerated in the jail
at some time before their diagnosis.
Intervention
When this outbreak was recognized, the local
health department worked with the jail to institute a
more thorough screening procedure during intake
and to encourage referral of inmates suspected of
having tuberculosis to the health department tuberculosis clinic for evaluation and treatment. The jail
also made efforts to improve their rates of tuberculin skin testing and is preparing to place ultraviolet
lights in high-traffic areas in the facility. To
promptly recognize potential clusters of related
cases, the health department now routinely does
restriction fragment length polymorphism testing of
isolates from persons with tuberculosis who have
recently been incarcerated.
19 October 1999
•
Annals of Internal Medicine
•
Volume 131
•
Number 8
561
Fortunately, no cases of drug-resistant tuberculosis were reported from this jail during the 3-year
investigation period. In an era when susceptibility
testing of M. tuberculosis isolates is routine, multidrug resistance is a readily available marker that
captures attention and may lead to recognition of
outbreaks in defined populations (such as prisoners). This investigation suggests that transmission of
drug-susceptible tuberculosis may occur in correctional facilities for prolonged periods before recognition, for lack of a clear marker of relatedness
between cases.
The incidence of tuberculosis in inmates during
the 3-year study period was 274 per 100 000 persons. In striking contrast, the rate of active tuberculosis in the United States in 1996 was 8 per
100 000 persons; in Tennessee, the rate was 10 per
100 000 persons, and in Memphis, the rate was 13
per 100 000 persons (1). In 1996, black men 25 to
44 years of age in the United States had a case rate
of 41 per 100 000 persons (1), and black men in this
age group in Memphis had a tuberculosis incidence
of 61 per 100 000 persons, less than one fourth of
the rate in the jail.
This analysis includes only inmates in whom tuberculosis was diagnosed while they were incarcerated or within 3 weeks of discharge from the jail.
Although the group includes persons who acquired
tuberculosis before entering the facility, many others who may have been infected in the jail but were
not given a diagnosis until after release are excluded. The number of persons exposed in the jail
but given a diagnosis months or years later cannot
be determined but could be substantial.
Our investigation highlights the difficulties in investigating and controlling tuberculosis in jails. Despite a strong epidemiologic link of cases with the
jail, even an extensive, retrospective analysis of
records could not identify specific person-to-person
lines of transmission among most of the inmates. As
demonstrated in Figure 4, overlapping periods of
incarceration could explain transmission among numerous inmates and guards infected with the same
strain of M. tuberculosis. Even in the confines of a
jail, high turnover and frequent movement of inmates provide opportunities for contacts that cannot
be completely elucidated by routine investigation.
Urban jails housing large numbers of inmates
may play a substantial role in the epidemiology of
tuberculosis in the surrounding community. Jail
populations have a high prevalence of risk factors
for communicable diseases. The potential for transfer of tuberculosis both into and out of such facilities is great. A large proportion (43%) of persons
with tuberculosis in Memphis have had previous
contact with the jail; although this does not imply
that their disease was a result of their incarceration,
562
19 October 1999
•
Annals of Internal Medicine
•
it does suggest that jails may provide a unique opportunity to screen and treat a population at high
risk that might not otherwise have access to medical
care. This could help limit the disease in the community as well as in the jail.
Such interventions would require a substantial
shift in priorities and application of resources. Correctional facilities have a mandate to safely separate
inmates from society and have not traditionally
been considered key participants in public health
interventions. Successful models that integrate public health systems and health care in correctional
facilities are being developed (26).
Screening inmates for active tuberculosis at the
time of admission and avoiding introduction of infectious persons into the jail population are critical
to the prevention of the spread of disease within
jails. Many prisons have programs that include placing a tuberculin skin test on all incoming inmates
and repeating the tuberculin skin test annually. Prisons have a relative advantage in that inmates will be
in the facility for extended periods. In contrast, jails
present unique logistic obstacles to screening; many
inmates are processed in and out daily, and many
stay for very short periods. While inmates are in the
jail, they are moved frequently and are often in
court or difficult to find, making the reading of
tuberculin skin tests challenging at best. This difficulty is evidenced by the low number of inmates
screened successfully with tuberculin skin tests in
this investigation and others (10, 27). Furthermore,
the sensitivity of the tuberculin skin test in screening for active tuberculosis is not high; 10% to 25%
of persons with active disease will have a negative
skin test result (28).
Screening for symptoms of tuberculosis is recommended as the first line of defense in situations with
a high prevalence of tuberculous disease (9, 28). All
inmates should have a systematic interview to determine the presence of symptoms of active disease
and whether they have had tuberculosis before or
have been told to take medications for this disease.
If symptoms such as cough lasting longer than 2
weeks, hemoptysis, chest pain, weight loss, fever, or
night sweats are present, inmates should be isolated
pending prompt evaluation. Screening inmates for
active tuberculosis with chest radiographs may also
be appropriate in jails with a high prevalence of
disease or risk factors for tuberculosis among inmates (9, 10, 27–29). The availability of new highspeed, low-dose miniature chest radiograph technology makes this a safer and more practical option
than in the past. Several jail systems are using universal chest radiograph screening for all jail admissions (30).
Routine cross-matching of incoming inmates with
local tuberculosis registries may also be useful in
Volume 131
•
Number 8
some settings to identify persons with histories of
incompletely treated disease or other risk factors
requiring follow-up. In addition, jail staff should be
educated about signs and symptoms of tuberculosis
and encouraged to facilitate prompt evaluation of
inmates suspected of having the disease. Reduction
of crowding (31), appropriate use of preventive
therapy, and use of high-efficiency particulate air
filtration and ultraviolet germicidal irradiation in
high-risk areas can also help control tuberculosis in
correctional facilities (9, 28, 32).
Control of tuberculosis in jails and prisons should
be an integral part of interventions directed toward
eliminating the disease in a community. Cooperation between jail medical staff and public health
tuberculosis programs is critical in ensuring appropriate treatment and follow-up of patients as they
move between correctional facilities and the community.
2. Hutton MD, Cauthen GM, Bloch AB. Results of a 29-state survey of
tuberculosis in nursing homes and correctional facilities. Public Health Rep.
1993;108:305-14.
3. Valway SE, Greifinger RB, Papania M, Kilburn JO, Woodley C, DiFerdinando GT, et al. Multidrug-resistant tuberculosis in the New York State
prison system, 1990-1991. J Infect Dis. 1994;170:151-6.
4. Braun MM, Truman BI, Maguire B, DiFerdinando GT Jr, Wormser G,
Broaddus R, et al. Increasing incidence of tuberculosis in a prison inmate
population. Association with HIV infection. JAMA. 1989;261:393-7.
5. Stead WW. Undetected tuberculosis in prison. Source of infection for community at large. JAMA. 1978;240:2544-7.
6. Bergmire-Sweat D, Barnett BJ, Harris SL, Taylor JP, Mazurek GH,
Reddy V. Tuberculosis outbreak in a Texas prison, 1994. Epidemiol Infect.
1996;117:485-92.
7. Koo DT, Baron RC, Rutherford GW. Transmission of Mycobacterium tuberculosis in a California state prison. Am J Public Health. 1991;87:279-82.
8. Fountain FF Jr. Tuberculosis in Shelby County and Tennessee correctional
facilities. Tenn Med. 1997;90:138-40.
9. Prevention and control of tuberculosis in correctional facilities: recommendations of the Advisory Council for the Elimination of Tuberculosis. MMWR
Morb Mortal Wkly Rep. 1996;45(RR-8):1-26.
10. Skolnick AA. Correction facility TB rates soar; some jails bring back chest
roentgenograms [News]. JAMA. 1992;268:3175-6.
11. Abeles H, Feibes H, Mandel E, Girard JA. The large city prison—a reservoir of tuberculosis. Tuberculosis control among sentenced male prisoners in
New York City. Am Rev Respir Dis. 1970;101:706-9.
12. Snider DE Jr, Hutton MD. Tuberculosis in correctional institutions. JAMA.
1989;261:436-7.
13. Bellin EY, Fletcher DD, Safyer SM. Association of tuberculosis infection
with increased time in or admission to the New York City jail system. JAMA.
1993;269:2228-31.
14. Pelletier AR, DiFerdinando GT Jr, Greenberg AG, Sosin DM, Jones WD
Jr, Bloch AB, et al. Tuberculosis in a correctional facility. Arch Intern Med.
1993;153:2692-5.
15. Glaser JB, Greifinger RB. Correctional health care: a public health opportunity. Ann Intern Med. 1993;118:139-45.
16. Bureau of Justice Statistics. Jails and jail inmates 1993-1994; census of jails
and annual survey of jails. In: BJS Bulletin. Washington, DC: Department of
Justice; 1995.
17. Bureau of Justice Statistics. Profile of Jail Inmates 1996. Washington, DC:
Department of Justice, Bureau of Justice Statistics, National Center for Justice
Statistics; 1998.
18. Tuberculosis transmission in a state correctional institution—California, 19901991. MMWR Morb Mortal Wkly Rep. 1992;41:927-9.
19. Valway SE, Richards SB, Kovacovich J, Greifinger RB, Crawford JT,
Dooley SW. Outbreak of multi-drug-resistant tuberculosis in a New York
State prison, 1991. Am J Epidemiol. 1994;140:113-22.
20. King L, Geis G. Tuberculosis transmission in a large urban jail. JAMA. 1977;
237:791-2.
21. Case definitions for infectious conditions under public health surveillance.
MMWR Morb Mortal Wkly Rep. 1997;46(RR-10):1-55.
22. Cave MD, Eisenach KD, McDermott PF, Bates JH, Crawford JT. IS6110:
conservation of sequence in the Mycobacterium tuberculosis complex and its
utilization in DNA fingerprinting. Mol Cell Probes. 1991;5:73-80.
23. Dean AD, Dean JQ, Coulombier D, Brendel KA, Smith DC, Burton AH,
et al. Epi Info, Version 6: A Word Processing, Database and Statistics Program
for Epidemiology on Microcomputers. Atlanta: Centers for Disease Control
and Prevention; 1994.
24. Braden CR, Templeton GL, Cave MD, Valway S, Onorato IM, Castro
KG, et al. Interpretation of restriction fragment length polymorphism analysis
of Mycobacterium tuberculosis isolates from a state with a large rural population. J Infect Dis. 1997;175:1446-52.
25. Sreevatsan S, Pan X, Stockbauer KE, Connell ND, Kreiswirth BN, Whittam TS, et al. Restricted structural gene polymorphism in the Mycobacterium tuberculosis complex indicates evolutionarily recent global dissemination.
Proc Nat Acad Sci U S A. 1997;94:9869-74.
26. Conklin TJ, Lincoln T, Flanigan TP. A public health model to connect
correctional health care with communities. Am J Public Health. 1998;88:
1249-50.
27. Puissis M, Feinglass J, Lidow E, Mansour M. Radiographic screening for
tuberculosis in a large urban county jail. Public Health Rep. 1996;111:330-4.
28. Controlling TB in Correctional Facilities. Atlanta: Centers for Disease Control
and Prevention; 1995.
29. Layton MC, Henning KJ, Alexander TA, Gooding AL, Reid C, Heman
BM, et al. Universal radiographic screening for tuberculosis among inmates
upon admission to jail. Am J Public Health. 1997;87:1335-7.
30. Wilcock K, Hammett TM, Widom R, Epstein J. Tuberculosis in correctional
facilities, 1994-1995. Washington, DC: Department of Justice, Office of Justice
Programs, National Institute of Justice; 1996.
31. MacIntyre CR, Kendig N, Kummer L, Birago S, Graham NM. Impact of
tuberculosis control measures and crowding on the incidence of tuberculous
infection in Maryland prisons. Clin Infect Dis. 1997;24:1060-7.
32. Hammett TM, Harvey S. Tuberculosis in Correctional Facilities. Washington,
DC: Department of Justice, Office of Justice Programs, National Institute of
Justice; 1994.
Presented in part at the 36th Annual Meeting of the Infectious
Diseases Society of America, Denver, Colorado, 12 to 15 November 1998.
From the Centers for Disease Control and Prevention, Atlanta,
Georgia; and the Tennessee Department of Health and Vanderbilt University School of Medicine, Nashville, Tennessee.
Acknowledgments: The authors thank Donna Hurst and William
L. Moore, MD, of the Tennessee Department of Health; David
Crowder and Gail Grant at the Centers for Disease Control and
Prevention; Yvonne Madlock, Martha Woloshyn, RN, Linda
Hamer, RN, Harold Brown, Francis Fountain, MD, and the staffs
of the Memphis/Shelby County Health Department, Correctional
Medical Services, Inc., the Memphis/Shelby County Criminal Justice Center, and the Shelby County Correctional Center for
assistance with the investigation; Dr. James Musser, Baylor College of Medicine, for genetic polymorphism testing; and Kenneth
G. Castro, MD, and Laura J. Fehrs, MD, Centers for Disease
Control and Prevention, for their thoughtful review of the manuscript.
Requests for Reprints: Timothy F. Jones, MD, Tennessee Department of Health, CEDS, Cordell Hull Building, 4th Floor, 425 5th
Avenue North, Nashville, TN 37247-5281. For reprint orders in
quantities exceeding 100, please contact Barbara Hudson, Reprints Coordinator; phone, 215-351-2657; e-mail, bhudson@mail
.acponline.org.
Current Author Addresses: Drs. Jones and Craig: Tennessee Department of Health, CEDS, Cordell Hull Building, 4th Floor, 425
5th Avenue North, Nashville, TN 37247-5281.
Dr. Valway: Division of Tuberculosis Elimination, National Center for HIV/STD/TB Prevention, Centers for Disease Control
and Prevention, Mailstop E-10, 1600 Clifton Road, NE, Atlanta,
GA 30333.
Dr. Woodley: Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Mailstop F-08, 1600 Clifton Road, NE,
Atlanta, GA 30333.
Dr. Schaffner: Departments of Medicine and Preventive Medicine, A1124 Medical Center North, Vanderbilt University School
of Medicine, Nashville, TN 37232-2637.
References
1. Reported Tuberculosis in the United States. Atlanta: Centers for Disease Control and Prevention; 1997.
19 October 1999
•
Annals of Internal Medicine
•
Volume 131
•
Number 8
563