Epidemiology of tuberculosis among the foreign-born in the United States
Mailman School of Public Health
April 7, 2004

	Amy Davidow, Ph.D.
	Asst. Professor of Preventive Medicine
	& Community Health
	Member, NJMS National Tuberculosis Center
	New Jersey Medical School
	Newark, NJ

Overview

	The problem
	Methods of approach; strengths & weaknesses
		Surveillance data
		Molecular epidemiology
	Where do we go from here?

WHO: 1/3 of the world has latent tuberculosis infection (LTBI)

WHO high-burden TB countries, 2004 (>80% of global TB)

	Afghanistan
	Bangladesh
	Brazil
	Cambodia
	China
	Democratic Rep. of Congo
	Ethiopia
	India
	Indonesia
	Kenya
	Mozambique
	Myanmar
	Nigeria
	Pakistan
	Philippines
	Russian Federation
	South Africa
	Thailand
	Uganda
	Tanzania
	Viet Nam
	Zimbabwe

Slide 5

Slide 6

Slide 7

Slide 8

Slide 9

We are not alone

	What is happening in US has happened/is happening elsewhere:
	When did foreign-born TB cases exceed 50% of reported cases in other countries?
		France:  1985
		Canada: 1990
		Netherlands:  1996
		US:  2003

TB in established market countries

	US, Canada, Western Europe, Israel, Australia, New Zealand, Japan
	Comparisons can be difficult
		Various definitions of foreign-birth:  country of birth, country of citizenship, ethnicity
		Country of origin may be missing by design (illegal to collect)

Slide 12

Slide 13

Israel: dramatic changes in a low prevalence country

	1989-95:  Population grew  by 1 Million
		2002 Population = 6.1 Million
		Europe/America-born 32.1%, Africa-born 14.6%, Asia-born 12.6% (2002)  (from CIA Factbook)
	4-fold increase in TB 1989-91 (Chemtob, 2002 & 2003)
		FB TB 80-85% of all TB
		former Soviet Union (>25% of cases in 1996):  38-172 per 100K
		Ethiopia (54% of cases in 1991):  500-3000 per 100K

Surveillance Studies

	What can we learn from them?

CDC studies of registry data (1)

	McKenna MT, McCray E, Onorato I. The epidemiology of TB among foreign-born persons in the US, 1986-1993. (NEJM 1995).
		55% of cases diagnosed < 5 yrs; 30% < 1 yr post-arrival
		More cases in younger immigrants than older immigrants, but lower case rate:  cohort effect?
		Largest relative difference between US-born and FB TB rates is among aged <15 yrs
		è substantial recent transmission around time of immigration (pre and post)

CDC studies of registry data (2)

	Zuber PT, McKenna MT, et al.  Long-term risk of tuberculosis among foreign-born persons in the United States. (JAMA 1997)
		Long term residents arriving aged > 5 yrs have TB rate 2-6 times the rate of those who arrived before their 5th birthday èImported TB responsible for most FB TB
		Selective screening needs to be adapted to local circumstances – places of origin, SES, migration patterns

Drug resistance and the foreign-born TB case

	More complicated and expensive to treat
	Association with time in US
			Greater rate among recent arrivals
			TB acquired in country of origin?
	Rx for LTBI among FB needed, esp. those from high prevalence countries, but may be inefficacious if there is resistance

Slide 19

CDC studies of registry data (3)

	Talbot EA, Moore M, et al. TB among foreign-born persons in the US, 1993-98. (JAMA 2000)
		CA, NY, TX, FL, NJ, IL = 73.4% of FB TB
		Most common birth countries vary by state:
			TX, CA, IL:  Mexico;  FL:  Haiti;  NJ:  India;  NY:  China, Dominican Republic, Haiti
		10% have known HIV infection
			less likely to be paired with TB as HIV infection is excludable condition for entry to US
			More than half of FB HIV/TB is in CA or NY
			Mostly among persons from Haiti or Mexico

CDC studies of registry data
(3, continued)

	Diagnosis of pulmonary TB in FB more likely by clinical criteria than in US-born:  14.3% vs. 10.9%
			FB more likely than US-born to be smear-negative
				47.3% vs. 36.7%
			And more likely to be culture-negative
				17.4 vs. 12.2%
		High index of suspicion for TB among FB when chest radiograph is abnormal OR
		Incomplete treatment prior to immigration?

CDC studies of registry data
(3, continued)

	TB control activities targeting prompt identification of TB and completion of therapy will not reduce TB among the FB
	Geographic variation of TB requires locally tailored approaches
		Areas with recent (case identification) vs. remote arrivals (screen for LTBI)
		Areas of high isonaizid resistance may require alternative LTBI treatment regimens

Surveillance cannot tell us (1)

	Are persons with active disease entering the US?
		Screening of immigrants – does it work?
		Contribution of non-screened foreign-born
			Temporary workers
			International students
			Undocumented
	Is current transmission taking place in the US?
		Within foreign-born communities
		From/to the foreign-born to/from the US-born

Surveillance cannot tell us (2)

	Among FB persons with latent TB infection (LTBI), who are high risk groups, i.e., likely to develop active TB?
	Who will accept treatment for LTBI?  Who will complete treatment?

Surveillance cannot tell us (3)

	How a patient’s lack of understanding of TB, cultural misunderstandings, economic barriers, lack of acculturation, etc. can contribute to delays in diagnosis
	How the health care system and health care providers can contribute to delays in diagnosis

Are persons with active disease entering the US?

Screening of immigrants as a TB control activity

	Who is screened?
		Screened persons are those applying for permanent residence (overseas or in US) or refugee status
		Immigration & Control Act of 1986: undocumented regularize status
	Classifications
		Active, smear positive TB cases – excludable condition
		B notifications – reports sent to local health departments (HDs), immigrants told to report to HDs
			B1 chest radiograph suggesting active TB but negative sputum
			B2 chest radiograph compatible with inactive TB

Some follow-up studies of B notifications (1)

	DeRiemer K, Chin DP, et al. 1998
		893 immigrants & refugees with San Francisco as intended destination and a referral for further medical evaluation
		84% sought further medical evaluation
		7% had active TB:  Class B-1 predictor of TB:  3.5 OR

Studies of follow-up (2)

	Zuber PL, Knowles LS et al. 1996
		Los Angeles County registry matched against tracking system for immigrants & refugees with suspected TB
		Tracking system contained
			5% of Mexican and Central American cases
			48% of NE Asian cases (Chinese, Korea, etc.)
			67% of SE Asian cases (Viet Nam, Thailand, etc.)

Studies of follow-up (3)

	Sciortino S, Mohle-Boetani, et al.,1999

Sciortino S, Mohle-Boetani, et al.1999 (continued)

	But B notifications did not identify 87% of the smear-positive adult TB cases!

Screening of international students - NO

	500,000 + international students in the US in 2000-2001.
		Top 5 countries:  India, China, Korea, Japan, Taiwan (Institute of International Education)
	CDC (Hennessey KA, 1998):  screening for LTBI among college students is inconsistent and problematic
	Texas (Weis SE, 2001), Ohio (Nelson ME , 1995):  TB among non-screened visitors is substantial

Screening of temporary workers - NO

	MMWR 45(47):1032-6, 1996.
		181 FB Hispanic TB patients in eight US counties in AZ, NM, TX, CA bordering Mexico, 1995.
			169 interviewed for the study, visa status not collected
		82% returned at least once to their country of origin
			35% returned at least monthly in the year preceding diagnosis
	Migrant workers
		Difficulties in treating mobile populations
		Migrant Clinicians Network www.migrantclinician.org
			Restricted circuit, point-to-point, nomadic

H-1B visa category

	For professionals working in specialty occupations; limited to 65,000 annually
	Created by Immigration Act of 1990
		Pre-1990:  Abnormal x-rays plus negative sputum required waivers to enter country
		Post-1990:  Liberalization:  to discourage sub-optimal overseas treatment
	Incidence of TB?  Unknown.

Census 2000 estimates of temporary workers by selected countries of origin

Is current transmission taking place in the US?

	Within foreign-born communities
	From/to the foreign-born to/from the US-born

Molecular epidemiology (1)

	Identical fingerprints thought to represent recently transmitted disease (Alland et al.  Bronx, NY &  Small et al.  San Francisco, NEJM 1994)
	US-born more likely than FB to have clustered (identical) IS6110 fingerprints
	Lack of fingerprint clustering among FB means reactivation, yet surveillance studies point to recently acquired disease!
		Catchment area:  FB from particular country/region in US.  What about the those remaining back home?

Molecular epidemiology (2)

	Secondary typing methods
		reduce extent of clustering (Burman WJ, 1997)
			è reduce the proportion of TB due to “recent infection”
		Validation: using epidemiologic links
		Links found for
			11% of patients with discordant fingerprints
			78% of patient isolates that matched by both IS6110 and pTBN12

Molecular epidemiology (3)

	BUT there is clustering among FB TB
		El Sahly et al., 2001:  30% of FB TB in Houston
		Ellis BA et al., 2002:  35% of FB TB
			AR, MD, MS, MI, NJ, Dallas plus 3 Counties in TX; and 6 Counties in CA
	Recent transmission?
	Limited genetic diversity in the country of origin (founder effect)?

Molecular epidemiology (4)

	Is transmission from the foreign-born to non-foreign-born occurring?
		San Francisco:  In 8 of 9 clusters that included both US & Mexican-born, index case was US-born (Jasmer RM et al., 1997)
		Netherlands:  RFLP shows transmission within FB communities and from FB to Dutch (Borgdorff et al., 1998)

Among FB persons with LTBI, who are high risk groups?

	Especially high-risk:  children, health care personnel, the HIV infected, people with other co-morbidities (diabetes), smokers (?)
	Who will accept treatment for LTBI?  Who will complete treatment?

“Foreign-born” children

	Higher prevalence of LTBI among children with FB parents, visitors from abroad, travel abroad (Lobato M et al., 1998)
	Source cases: < 50% of children have one
		Harder to identify for FB children
		However, of children with potential source cases, >50% of the source cases are FB  (Sun SJ et al., 2002)

Occupational health

	FB health care personnel
		hard to interpret annual TST:  BCG?  LTBI acquired in country of origin?
	FB TB patients more likely to be working than US-born TB patients
		Implications for workplace contact investigations
			Kim DY, Ridzon R, et al., 2002: DE poultry workers,  work-related cluster ruled out using spoligotyping
		Undocumented workers in particular industries

Where does surveillance
go from here?

	RVCT Revision Working Group
		projected roll-out 2006
		Last revision 1992
	TB Epidemiologic Studies Consortium, Task 9
		“Enhanced surveillance to identify missed opportunities for prevention of tuberculosis in the foreign-born”
		pilot study beginning April 2004

Where does molecular epidemiology go from here?

	Many secondary typing methods available
		Spoligotyping, others
	Approaches to quantify the extent to which  fingerprints do not match
		Genetic distance: expected waiting time for the steps required to diverge from a hypothetical common ancestor
		Dice coefficient:  measure of similarity
	Is an identical fingerprint necessary to conclude that there is a recent chain of transmission?