Abstract
Objectives. This study sought to determine adherence of physicians to tuberculosis (TB) screening guidelines among foreign-born persons living in the United States who were applying for permanent residency.
Methods. Medical forms of applicants from 5 geographic areas were reviewed, along with information from a national physician database on attending physicians. Applicant and corresponding physician characteristics were compared among those who were and were not correctly screened.
Results. Of 5739 applicants eligible for screening via tuberculin skin test, 75% were appropriately screened. Except in San Diego, where 11% of the applicants received no screening, most of the inappropriate screening resulted from the use of chest x-rays as the initial screening tool.
Conclusions. Focused physician education and periodic monitoring of adherence to screening guidelines are warranted.
Individuals born outside the United States accounted for 46% of the 16 377 US cases of tuberculosis (TB) reported in 2000.1 To eliminate TB in the United States, the Institute of Medicine and others have identified screening for the disease as a priority among individuals applying for permanent US residence.2–4 Previous studies have shown a high yield of cases from overseas screening5,6; however, with the exception of a limited evaluation in Denver, Colo,7 in which the health department served as the sole screening agency for mostly Mexican applicants, adherence to or yield from the US-based screening process has not been evaluated.
In the United States, through the process of applying for permanent US residency, 578 000 persons were screened in 1995, and this number increased to 760 000 in 1997.8 These applicants undergo TB screening, and adherence to and yield from the recommended screening should interest public health officials. Applicants 2 years or older are given a Mantoux tuberculin skin test (TST); if a reaction of 5 mm or more is observed, a chest x-ray is required. If the x-ray suggests TB, the applicant must be referred to the local health department for evaluation. The guidelines also recommend referring applicants with a TST induration of 10 mm or more and a chest x-ray not suggestive of TB to the health department for consideration of preventive therapy.9
Immigration and Naturalization Service (INS) district directors appoint individual physicians (known as civil surgeons) to perform screening examinations; these physicians must be licensed, have at least 4 years of experience, and be provided with guidelines (Technical Instructions for Medical Examination of Aliens in the United States9) developed by the Division of Quarantine of the Centers for Disease Control and Prevention (CDC). After performing examinations, physicians complete INS medical forms and return them to the INS. The United States has approximately 3000 civil surgeons, but the number currently performing examinations of applicants is unknown.
METHODS
We evaluated adherence to the CDC's technical instructions in 5 INS jurisdictions with large numbers of applicants for permanent US residence: San Diego County, California; San Francisco County, California; New York State (outside New York City); New York City; and Massachusetts. Most of the applicants were seeking to join immediate relatives who are US citizens.
Trained data abstractors recorded information from the INS medical form and any laboratory or radiology reports. Information included applicant's name, sex, date of birth, country of birth, and residence; name of attending civil surgeon; date of examination; and complete results of the physical examination, including all tests except HIV. Abstractors also recorded (1) whether a TST was performed and, if so, the size of the induration; (2) chest x-ray results; and (3) whether the physician considered the applicant to have active infectious TB as indicated by chest x-ray and microscopy results.
Annual number of applicant records processed ranged from 3000 in New York State to 22 000 in San Francisco and 94 000 in New York City (oral communication; C. Quemby, INS; September 1997). Records were selected by different sites in a systematic manner as follows: (1) sequentially for a specified period, (2) every nth record of the total number of records present, or (3) every nth day for specific intervals during the study period. Ten percent of the total records were validated by a data abstractor not involved in the original abstracting process. Files that did not contain a medical form were omitted, and the next file that contained a medical form was used.
We used the American Medical Association (AMA) Internet-based physician select service (written communication; P. L. Havlicek, PhD; June 1998)10 to obtain information on 4 characteristics of civil surgeons: medical school training (US vs foreign), years since medical school graduation, self-reported specialty (primary care vs other), and board certification status. Although accuracy is not guaranteed, information in this database has been verified and authenticated by accrediting agencies, medical schools, residency training programs, licensing and certifying boards, and other data sources. After obtaining CDC institutional review board approval, we linked physician characteristics to actual TB screening practices for each area.
We stratified all analyses by US residence of the applicant. Because the INS instructions differ from the CDC's technical instructions by stating that civil surgeons need to order TSTs only for individuals 15 years or older (rather than 2 years or older), and because at the time of the study only individuals younger than 35 years were generally considered candidates for preventive therapy,11 we used the following age groups: 2 to 14 years, 15 to 34 years, and 35 years or above. Countries of birth were classified into the following categories: Asia, Africa, Mexico and Central America, South America, the Caribbean, the Middle East, Eastern Europe, and “established market economies” (defined as Western Europe, Japan, Canada, Australia, New Zealand, and Israel). In determining the number of persons eligible for TST, we excluded applicants younger than 2 years as well as those with previous positive TST results, isoniazid preventive therapy, or TB disease.
Five TB screening outcomes were identified: (1) no TST or chest x-ray screening, (2) chest x-ray but no TST, (3) TST with an induration of 5 mm or more but no chest x-ray, (4) TST with an induration below 5 mm regardless of chest x-ray, and (5) both a TST with an induration below 5 mm and a chest x-ray. To identify applicant characteristics (age, sex, region of origin, whether a bacille Calmette–Guerin [BCG] scar was noted, number of applicants per physician) associated with nonadherence to the required initial TST, we compared applicants in each area who received no screening (outcome 1) or chest x-ray alone (outcome 2) with applicants who were correctly screened (outcomes 4 and 5). Because only 1% of the applicants experienced outcome 3, this group was excluded from analyses of nonadherence.
We used the same approach to identify physician characteristics (sex, medical school training, specialty, and board certification) associated with nonadherence to TST recommendations; however, the unit of analysis was examination. Epi Info version 6.04b (CDC, Atlanta, Ga) was used to enter information into a database, and SAS version 6.11 (SAS Institute Inc, Cary, NC) was used in conducting Pearson χ2 tests. A significance level of P < .05 was used in analyses.
RESULTS
After exclusion of 11 records that were incomplete, 5739 records remained for analysis (Table 1 ▶). The mean age of all applicants was 30.4 years; San Diego applicants were significantly younger than average (P < .0001). In all of the study areas, more than 60% of applicants were aged 2 to 34 years. Distributions by sex were generally similar except in San Diego, where more than 64% of applicants were female (P < .0001).
TABLE 1—
Demographic Characteristics of Applicants With Medical Forms by Study Area: 5 US Sites, 1997–1998
| Characteristic | San Diego (n = 1394) | San Francisco (n = 1099) | New York State (n = 1073) | New York City (n = 1157) | Massachusetts (n = 1016) |
| Mean agea, y | 25.9 | 32.5 | 30.9 | 32 | 32.1 |
| Age group, y, No. (%) | |||||
| 0–1 | 17 (1) | 2 (0) | 3 (0) | 0 (0) | 2 (0) |
| 2–14 | 244 (18) | 63 (6) | 84 (8) | 68 (6) | 56 (6) |
| 15–34 | 872 (63) | 702 (64) | 668 (62) | 680 (59) | 644 (63) |
| ≥35 | 261 (19) | 332 (30) | 318 (30) | 409 (35) | 314 (31) |
| Region of birth | |||||
| Asia | 149 (11) | 372 (34) | 163 (15) | 165 (14) | 77 (8) |
| Africa | 23 (2) | 17 (2) | 54 (5) | 89 (8) | 150 (15) |
| Central America and Mexico | 1022 (73) | 327 (30) | 151 (14) | 88 (8) | 101 (10) |
| South America | 20 (1) | 72 (7) | 152 (14) | 192 (17) | 156 (15) |
| Caribbean | 11 (1) | 7 (1) | 172 (16) | 297 (26) | 226 (22) |
| Middle East | 39 (3) | 38 (3) | 69 (6) | 100 (9) | 84 (8) |
| Eastern Europe | 18 (1) | 72 (7) | 118 (11) | 156 (13) | 65 (6) |
| Established market economies | 112 (8) | 194 (18) | 194 (18) | 70 (6) | 157 (15) |
Note. As a result of rounding, percentages may not sum to 100%.
aAge is normally distributed.
Region-of-birth distributions differed noticeably by study area. For example, almost three fourths (73%) of San Diego applicants were from Mexico or Central America, and 11% were from Asia; in San Francisco, however, 34% were from Asia, 30% were from Mexico or Central America, and 18% were from established market economy countries (Table 1 ▶).
San Diego included the highest percentage (11%) of applicants who were not screened at all, but most applicants with a TST reaction of 5 mm or more had the recommended chest x-ray (Table 2 ▶). In terms of TSTs administered, San Francisco exhibited the poorest performance (only 48% of applicants), but 52% of its applicants had a chest x-ray rather than a TST. New York City had the best record of adherence with the guidelines (91% of applicants underwent a TST), followed by Massachusetts (87%), New York State (75%), and San Diego (74%; Table 2 ▶).
TABLE 2—
TB Screening Outcomes of Applicants by Study Area: 5 US Sites, 1997–1998
| Outcome, No. (%) | San Diego (n = 1345) | San Francisco (n = 1074) | New York State (n = 1050) | New York City (n = 1141) | Massachusetts (n = 960) |
| No TST, no CXR | 153 (11) | 6 (1) | 40 (4) | 10 (1) | 12 (1) |
| CXR, no TST | 192 (14) | 555 (52) | 228 (22) | 96 (8) | 101 (11) |
| TST < 5 mm | 768 (57) | 385 (36) | 582 (55) | 891 (78) | 748 (78) |
| TST ≥ 5 mm, no CXR | 22 (2) | 9 (1) | 14 (1) | 7 (1) | 4 (0) |
| TST ≥ 5 mm, CXR | 210 (16) | 119 (11) | 186 (18) | 137 (12) | 95 (10) |
| TST ≥ 10 mma | 199 (20) | 98 (19) | 187 (24) | 122 (12) | 88 (10) |
Note. TST = tuberculin skin test; CXR = chest x-ray. As a result of rounding, percentages may not sum to 100%. Applicants younger than 2 years were excluded, as were others with a history of a positive TST, isoniazid therapy, or TB disease.
aPercentage of those with TST performed.
We identified several characteristics that were significantly more common among applicants with no screening or screening by chest x-ray than among applicants with TST screening. For example, in New York State and Massachusetts higher proportions of unscreened applicants were from Asia and countries with established market economies, and in New York State and San Diego more such applicants were in the 2- to 14-year age group (data not shown). In San Francisco and New York State, characteristics associated with a greater likelihood of screening via chest x-ray than via TST included older age (35 years or above), physician practices with higher proportions of Asian patients, and higher applicant volumes (data not shown).
Of the 491 physicians, 70% were found on the AMA Web listing (range: 62% in Massachusetts to 83% in San Diego). Except in New York City and New York State, most of these physicians had graduated from US medical schools; in all areas, more than half of the physicians (range: 53%–58%) had graduated more than 20 years before the study period. Seventy-five percent were primary care practitioners, 47% were board certified (range: 32% in San Diego to 73% in Massachusetts), and 81% were male (range: 70% in Massachusetts to 85% in New York City and New York State). Board certification, primary care, graduation from a US medical school, and being a recent graduate were predictive of nonadherence to guidelines at different sites, but no consistent pattern emerged (data not shown).
No active cases of TB were newly diagnosed as a result of screening (95% confidence interval [CI] = 0%, 0.0007%); 10 applicants had a history of current or past TB noted on INS forms. When TST positivity was defined as an induration of 10 mm or more, TST positivity rates ranged from 10% among Massachusetts applicants to 24% among New York State applicants.
DISCUSSION
Limitations of the study, including the fact that the AMA database did not contain information on 30% of the physicians, may have biased our results. We were also limited in terms of the applicant and physician variables we could examine. Furthermore, with the wide variability in practices between the study areas, the physicians may not have been representative of all physicians who serve in this capacity, or a few physicians with large applicant volumes may have skewed the results.
Even so, this study provides an important initial evaluation of the US-based screening process, and the findings can assist in the development of a working relationship involving key partners in immigration health: state and local health departments, the INS and CDC, and examining physicians. We found that 24% of applicants were not screened with TST, falling short of the Institute of Medicine recommendation.2 The lack of screening observed may be explained in part by the inconsistencies in testing age criteria between the INS form (15 years or above) and the CDC's technical instructions (2 years or above).
Why did the civil surgeons use a chest x-ray as the initial screening tool in 20% of the cases? In New York State, recording the presence of a BCG scar was associated with this approach (data not shown); some physicians may believe that BCG interferes with TST results. The reason for the association in several areas between screening by chest x-ray and Asian origin may have been that knowledge of TB prevalence in Asia influenced physicians' behavior.
Our study may have been too small to demonstrate the utility of screening in regard to case finding; for example, in comparison with the TB prevalence rate of 176 cases per 100 000 found in earlier research involving Vietnamese immigrants,12,13 our study size would have been expected to yield only 12 cases. However, given that 55% of TB cases among newly arrived immigrants are estimated to occur within the first 5 years of residence,13 it is likely that the US program would detect fewer cases than the overseas screening program. Most of those who apply for permanent residence have been in the United States for several years and may have already been treated for TB or may not have had ongoing exposure to the disease.
Although the process described identified relatively few TB cases, it still provides an excellent opportunity to detect candidates for preventive therapy. We found, however, that only 16% of applicants overall had TST indurations of 10 mm or more, and rates varied by study area. Among Mexican applicants, positivity rates ranged from 19% in San Diego and Massachusetts to 29% in New York City (data not shown), well below the 42% observed in the earlier-mentioned Denver study7; among Asians, rates ranged from 8% in New York City to 30% in San Diego (data not shown), again substantially lower than rates for Asian immigrants reported in other studies.14
Was there some systematic error in the screening process that lowered positivity rates? Several hypotheses might account for this observation: (1) the forms may be completed incorrectly, with many small reactions missed; (2) physicians may not be administering the TST correctly; (3) physicians may be improperly reading the results; (4) applicants may be self-reading their results; or (5) applicants may differ from other immigrants and refugees and may actually have a lower TST positivity rate.
As a result of this study's findings, efforts at both the national and state levels are being made to develop training materials that reflect established guidelines and recommendations for physicians who screen immigrants for TB infection and disease.2 These physicians need to be targeted to adopt appropriate screening practices so as to reduce the prevalence of TB among foreign-born Americans. As illustrated by the Denver study, this US-based screening process has a high ability to identify persons with TB infection and is cost-effective.7
Acknowledgments
We thank Joseph Foster, JD, of the Centers for Disease Control and Prevention, and the INS for contributing to the implementation of the study.
M. Saraiya designed the study, analyzed the data, and wrote all drafts of the paper. S. T. Cookson, P. Tribble, B. Silk, and N. J. Binkin made substantial contributions to the interpretation of the data and subsequent revisions of the manuscript. B. Silk, R. Cass, S. Poonja, M. Walting, N. Howland, E. A. Paz, J. Cochran, K. S. Moser, and M. J. Oxtoby provided administrative, technical, and logistical support and collected and assembled the data. All of the authors participated in interpreting the data and editing drafts of the manuscript.
Peer Reviewed
References
- 1.Centers for Disease Control and Prevention. Tuberculosis morbidity among US-born and foreign-born population—United States, 2000. MMWR Morb Mortal Wkly Rep. 2002; 51(05);101-4. [PubMed] [Google Scholar]
- 2.Institute of Medicine. Ending Neglect: The Elimination of Tuberculosis in the United States. Washington, DC: National Academy Press; 2000. [PubMed]
- 3.Centers for Disease Control and Prevention. Recommendations for prevention and control of tuberculosis among foreign-born persons. MMWR Morb Mortal Wkly Rep. 1998;47:1–26.9450721 [Google Scholar]
- 4.Saraiya M, Binkin NJ. Tuberculosis in immigrants. In: Reichman LB, Hershfield ES, eds. Tuberculosis: A Comprehensive International Approach. New York, NY: Marcel Dekker; 2000:661–692.
- 5.Binkin N, Zuber P, Wells C, Tipple M, Castro K. Overseas screening for tuberculosis in immigrants and refugees to the United States: current status. Clin Infect Dis. 1996;26:1226–1232. [DOI] [PubMed] [Google Scholar]
- 6.Wells C, Zuber P, Nolan C, Binkin N, Goldberg S. Tuberculosis prevention among foreign-born persons in Seattle–King County, Washington. Am J Respir Crit Care Med. 1997;156:573–577. [DOI] [PubMed] [Google Scholar]
- 7.Blum R, Polish L, Tapy J, Catlin B, Cohn D. Results of screening for tuberculosis in foreign-born persons applying for adjustment of immigration status. Chest. 1993;103:1670–1674. [DOI] [PubMed] [Google Scholar]
- 8.US Immigration and Naturalization Service. Legal immigration, fiscal year 1997. Available at: http://www.ins.usdoj.gov/graphics/publicaffairs/newsrels/97Legal.pdf. Accessed May 14, 2001.
- 9.Technical Instructions for Medical Examination of Aliens in the United States. Atlanta, Ga: Centers for Disease Control; 1991.
- 10.American Medical Association. AMA physician select. Available at: http://www.ama-assn.org/aps/amahg.htm. Accessed May 14, 2001.
- 11.Core Curriculum on Tuberculosis: What the Clinician Should Know. Atlanta, Ga: Centers for Disease Control and Prevention; 1994.
- 12.McKenna M, McCray E, Onorato I. The epidemiology of tuberculosis among foreign-born persons in the United States, 1986–1993. N Engl J Med. 1995;332:1071–1076. [DOI] [PubMed] [Google Scholar]
- 13.Zuber P, McKenna M, Binkin N, Onorato I, Castro K. Long-term risk of tuberculosis among foreign-born persons in the United States. JAMA. 1997;278:304–307. [PubMed] [Google Scholar]
- 14.Dye S, Scheele S, Dolin P, Patania V, Raviglione MC, for the WHO Global Surveillance and Monitoring Project. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. JAMA. 1999;287:677–686. [DOI] [PubMed] [Google Scholar]