The health of New York City’s immigrant population has profound implications for New York City’s economy and health sector. In 1990, almost 34% of New York City’s population was comprised of persons born outside of the United States or its territories.1 By 1999, the proportion of foreign-born persons residing in New York City had grown to almost 40%.2  Over this decade, the growth in the foreign-born population corresponded with substantial growth of the regional New York economy–an economy that is influenced by the productivity of the foreign-born to a larger extent than most other urban areas.3

A critical component of economic productivity is health.4-6 In addition, the New York regional economy depends heavily on the health sector.  Health care is the third largest economic sector in New York City, accounting for more than 13% of all employment and wages and is an economic sector heavily dependent on foreign-born workers and patients alike.7  The confluence of these circumstances, that is, the complex interaction between the demand and supply of services associated with foreign-born persons in the New York health sector and their health status, health outcomes, and economic productivity, is driving an important new momentum among health policy analysts to understand the health needs and health outcomes of the newest New Yorkers.

The national debate regarding the economic productivity and health status of the foreign-born is no less complex; the foreign-born are squarely in the middle of a national health policy controversy in the U.S. over whether health care sector spending is “out of control.”8  While no one debates that health is positively related to economic productivity, or that excess morbidity and mortality in a population have adverse economic consequences,9-14  there is little consensus over the effectiveness of increased health care expenditures in improving health.15   Recent ballot and legislative actions, including the 1994 passage of proposition 187 in California, and the 1996 passage of the Illegal Immigration Reform and Immigrant Responsibility Act, restrict access to publicly funded health and social services for both documented and undocumented immigrants.  Policy rationales for targeting foreign-born populations for health care benefit cuts may be influenced by the academic debate over health care effectiveness, as well as uncertainty surrounding returns on health investments for immigrant groups, given that they may be less healthy than native-born persons.  However, the most often discussed rationale behind these policy initiatives has been that immigrants are contributing to increased costs and draining our taxpayer resources by using more than their “share of health and social services.”16  Yet what is the validity of that rationale?

It is striking how little is known regarding how immigrants are affecting, or are affected by, the organization and delivery of health services.  Despite evidence to the contrary among some studies in the medical literature, the conventional wisdom in the United States today is that the foreign-born are sicker and in more need of health services than the native-born population.17  What is the reality?  Are immigrants sicker?  Do they cost the taxpayer more than native-born citizens?  Are their health profiles and service utilization similar or different from native-born groups? What are the determinants of health outcomes among different immigrant groups? There is a pressing need for systematic investigations of the health status, health needs, access to services, utilization, and treatment outcomes of recent immigrants to major urban centers such as New York City.

The purpose of this chapter is to contribute to the ongoing debate over immigrant health policy by beginning to systematically address some of the questions surrounding the health status of immigrants.  The particular focus of this investigation is to examine the health status of New York City’s foreign-born populations and their patterns of health service utilization. We begin with a descriptive analysis of basic health and health-related demographic information for foreign-born populations residing in New York City.  Following these results, we present an analysis of hospital utilization and mortality rates using currently available data from vital statistics, hospital admissions, Immigration and Naturalization Service, and the Census.  Finally, using the technique of “small area analysis,” 18 we undertake an empirical investigation of the determinants of hospital utililization among immigrants, examining the roles that income, gender, race, ethnicity, years of residence in the United States, housing conditions, and citizenship play among foreign-born and native-born persons.  In particular, we use multivariate techniques to examine and test the validity of the hypothesis that immigrants are “draining” health resources in New York City.

Background

Health Status

The health of foreign-born populations has long been a subject of considerable political controversy.  Historically, recent immigrants to the U.S. have been blamed for disease outbreaks.  Indeed, immigrant groups have often become metaphors for the stigma of disease, rather than pathogenic vectors, social conditions, or environmental factors.19-21 In the 19th century for instance, Irish immigrants were believed to be “inherently diseased” and were blamed for the first with outbreaks of cholera, and later typhoid.22  Italian immigrants were held responsible for the polio epidemic in Philadelphia, and Chinese in San Francisco blamed for the 1890 bubonic plague outbreak.  In the 1980’s, Haitians were singled out by the Centers for Disease Control as high risk for HIV/AIDS.23

More recently, the perception that immigrants are “sicker” than native-born populations has been challenged by the data.  The National Health Interview Survey began to include data on birthplace in 1985, and on length of residence in the United States in 1989.  Using these data, and controlling for socioeconomic characteristics, Stephens found that immigrants rank higher than native-born groups on all measures of health status, including functional, mental, and physical health status.  Recent immigrants were also found to rank higher in health status measurement scales than immigrants who have resided in the U.S. ten years or more.24 Moreover, in the first systematic study of the costs of illegal and legal immigration, including Medicaid costs, the Urban Institute found the total immigrant population is an economic boon, not a burden, for the country over all, generating a surplus of $25 billion to $30 billion.25

It is surprising that so little is known regarding the health status of foreign-born persons living in New York City.  With a few notable exceptions, what we know about immigrant health is derived from national data. Fairly consistent reports of good health outcomes among immigrant and other foreign-born populations have been reported in the medical literature.26 Foreign-born persons appear to have lower rates of chronic disease than do native-born persons.27  For instance, persons born in Asia, Latin America, and Africa are reported to have substantially lower age-adjusted cancer rates and heart disease rates than native-born Americans have.28 One study conducted in Canada indicates that healthcare utilization rates may be lower and life expectancy higher among legal immigrants to Canada than native-born Canadians.29

These studies have led to the “healthy migrant” hypothesis, which has been applied to immigrants and migrants alike.  This hypothesis suggests that immigrants and other foreign-born populations are a self-selected healthy population, due to the necessity of health for undertaking strenuous travel, meeting the physical and emotional challenges of migration, and earning a living in a new country.  Moreover, the United States and other developed nations attempt to exclude some people from legally immigrating if they have certain medical conditions.

Refugees may have different health risks than other foreign-born populations, and may be at greater risk for some diseases than their United States born counterparts.  Refugees, particularly from developing countries, may suffer from disproportionate rates of tuberculosis, parasitic disease, anemia, hepatitis B, malnutrition, poor immunization status, infectious skin disease, poor dentition, lead poisoning, and smoking.30-43  Rates of some infectious diseases exceed 50 percent among refugee groups and often include conditions North American physicians are not familiar with.44,45   The vast majority of these conditions, however, only rarely lead to hospitalization or death.

In addition, evidence from the domestic and international medical literature suggests that there is an increased incidence of depression and post-traumatic stress disorder in refugees from war torn regions, compared with the mental health status of non-refugee immigrants.38-48  However, while such problems are intuitive, questions surrounding even these data remain.  For example, survey data in one mental health facility indicates that immigrants may be reluctant to seek care for mental illness.49

Though New York State provides medical coverage to documented immigrants, and New York hospitals are prohibited from turning away emergency cases, federal law limits eligibility for federally funded medical care programs to persons who immigrated prior to 1996 or who became disabled while residing in the United States.50  Those who legally immigrated after 1996 may receive government benefits after five years of residence in the United States. The majority of undocumented immigrants do not have health insurance and may be reluctant to seek medical care for fear of deportation.51,52   Given these real and perceived barriers to health care access for immigrants, it is possible that some of the “healthy migrant effect” observed in the medical literature is attributable to immigrants’ reluctance to seek care, causing them to be uncounted in retrospective case-controlled studies or studies using claims-based databases.
In conclusion, despite popular perceptions of the stigma of disease associated with immigrant groups, there is an emerging consensus in the scientific literature that immigrant groups are often healthier, with some exceptions, than their native-born counterparts.  However, this tentative conclusion must be made with great caution.  Few scientific community-based epidemiologic surveys exist, due in part to their high expense.  Most of the findings from the literature are based on health service utilization statistics.  However, access to care may be limited by perceived or real barriers, such as lack of eligibility to some sources of public funding, lower rates of employer-based commercial insurance, and lack of culturally competent ambulatory care facilities.53

Health-Related Demographics

A large literature exists on the determinants of health. Known predictors of health status include age, income, education, overcrowded housing conditions, marital status, gender, race, ethnicity, access to medical care, substance abuse (including tobacco and alcohol), diet, and exercise.54-57 Income and educational attainment appear to be second only to natural human aging as determinants of the health of a population.

Over the past decade, research on the determinants of health has focused on socioeconomic, race and ethnic disparities in health service use and outcomes. Some of the most glaring differences in outcomes have been found among black and Hispanic minorities.  Infant mortality rates are nearly 2 times greater for blacks than whites.  The prevalence of diabetes in Hispanic Americans is approximately double that in whites.58  African American men living in Harlem have the same life expectancy as do men in Bangladesh.59  Black women die from breast cancer at 2½ times the rate of whites.60  These disparities persist despite improvements in health for the Nation as a whole.61  For example, mortality from ischemic heart disease decreased 20% for the overall population but only 13% for blacks between 1987 and 1995.62  As minority populations continue to increase in size and proportion, the health of these groups will increasingly affect the health and quality of life of New York City.63

Some argue that elimination of financial differences in access would profoundly reduce health disparities.64  Nevertheless, disparities exist within health systems.65  In a study of Medicare beneficiaries, providers gave less intensive treatment to blacks than to whites after controlling for other factors.66  A study of black and white males treated for lung carcinoma found that all-cause mortality was higher for blacks than whites due in part to differences in rates of surgery, chemotherapy, and radiation.67  A recent study of how physicians manage chest pain suggests that decision-making may be influenced by race and ethnicity after controlling for other factors.68

However, almost without exception, studies in the literature do not distinguish immigration status as an explanatory factor in either the conceptual or the analytic models.  This lack of distinction clouds our understanding of the causes and consequences of, and effective solutions for, these persistent disparities in health outcomes. It is possible for instance, that if the “healthy migrant effect” holds true, studies aggregating foreign-born and native born-minorities into one group defined by race, under-estimate health disparities among native-born minorities, an alarming possibility.

A rational public discourse is necessary regarding the demand for, and supply of, effective health services to maintain and improve the health of recent immigrants and their families.  The remaining sections of this chapter provide, to our knowledge, the first quantitative study of immigrant health and utilization patterns in NYC.  The analyses are not definitive but offer a preliminary assessment of explicit health and health care utilization patterns among immigrants in New York City in an attempt to make a positive contribution towards improving immigrant health outcomes in New York.

Research Methods

Overview and definitions

To ascertain the health status of foreign-born persons residing in New York City, we conducted two distinct investigations.  In the first, we examined computer-based hospitalization records to determine whether foreign-born persons are hospitalized at a higher or lower rate than native-born persons or persons born in Puerto Rico, a U.S. territory.  In this ecological analysis, we compared hospitalization rates in 55 neighborhoods in New York City using multivariable linear regression models.  After controlling for known risk factors, such as total household income, gender, and race, we examined whether the relative proportion of foreign-born persons residing in a given neighborhood might predict hospitalization rates for that neighborhood. In other words, we would expect that, if foreign-born persons had lower hospitalization rates than native-born persons, neighborhoods with high proportions of foreign-born residents would have lower overall hospitalization rates than neighborhoods with high concentrations of native-born persons.

For the purposes of this chapter, we use hospitalization data not only as a measure of health care utilization, but also as an indicator of health status.  Higher rates of hospitalization in predominately foreign-born neighborhoods would suggest that the foreign-born are in poor health compared with their U.S. born counterparts.  Yet the converse is not necessarily true.  Because hospitalization is required only for severe illness, a finding of lower hospitalization rates in predominantly foreign-born neighborhoods (relative to predominantly native-born neighborhoods) might be primarily due to a healthier foreign-born population.  However, given the widespread perception among the foreign-born that medical services are not available to them in the United States and real barriers to elective or other non-emergent hospital-based treatments exist, it is possible that lower hospitalization rates in foreign-born neighborhoods merely reflects underutilization of needed hospital services.

However, if the overall health status of foreign-born persons is worse than native-born persons and such persons are less likely to access care, we would expect higher mortality rates in this population.  In the second analysis, we examined age-specific mortality rates, age-adjusted mortality rates, and life expectancy for foreign-born persons, native-born persons, and persons born in Puerto Rico.  Because mortality is a sentinel indicator of the health status of a population and because these data are not subject to ecological confounders, mortality data perhaps offer the clearest picture of the health status of a population.

However, like hospitalization data, mortality data do not provide information on the utilization of ambulatory medical resources and do not provide a complete picture of the health status of the foreign-born population.  For example, if we know that mortality rates are higher among foreign-born persons than among native-born persons, we are left wondering whether this was due to poor health status or lack of access to ambulatory medical care.  Furthermore, if mortality rates are lower among foreign-born persons, the possibility exists that some persons returned to their country of birth to die.

Comparing hospitalization rates with mortality rates provides a more complete picture of the health status of foreign-born populations.  For example, a finding of relatively low hospitalization rates coupled with relatively high mortality rates among the foreign-born suggests that such persons are not accessing health services and are dying outside of the hospital.  Because ambulatory data are not available by neighborhood, as hospitalization data are, further investigations into these questions would require primary data collection.

While the term “immigrant” refers to persons choosing to permanently reside in the United States, the term “foreign-born” refers to persons residing in the United States at any point in time.  The analyses presented below are of all foreign-persons, including temporary workers and full-time residents regardless of whether their presence in the United States is documented or undocumented.  Persons born outside of the United States or its territories to parents who were U.S. citizens were not included in the foreign-born category.  Persons born in the continental United States and Hawaii were considered native-born and persons born in U.S. territories were analyzed separately.  Because virtually all of those persons born in U.S. territories who currently reside in New York were born in Puerto Rico, it was not possible to obtain a reliable estimate of persons born in other territories (e.g. the Virgin Islands) and these persons were excluded from the analysis entirely.

Data Sources

In this section, we will discuss the data that were used in our analysis of the health-related demographics of our cohort, our analysis of hospitalization rates, and our analysis of mortality rates among foreign-born, native-born, and Puerto Rican-born persons residing in New York City.

Hospital care utilization and morbidity analysis

We obtained population data, which included country of origin, age, income, gender, race, and ethnicity from the 1996 Housing and Vacancy Survey (HVS)–a survey of households conducted every three years by the United States Bureau of the Census.  We also obtained recent demographic data from the 1999 HVS, which added questions pertaining to the year of entry and provided more recent background information on the foreign-born population.   The 1996 HVS was used in the analysis of hospitalization rates rather than the 1999 HVS because the 1999 hospital datafile was not yet publicly available.

The HVS contains a sub-sample of 10,000 New York City households.  Country of origin is asked only of the householder and the householder’s parents.  Country of origin data are calculated under the assumption that other household members will be from the same country as the householder.  The survey was only administered to persons 18 years and over and sample weights were not available for geographic regions smaller than “sub-borough areas,” neighborhoods that consist of multiple census tracts.

Overall and diagnosis-specific hospitalization rates were calculated using 1996 hospital discharge data from the Statewide Planning and Research Cooperative System (SPARCS) in the numerator and 1996 HVS data in the denominator.  Although some coding errors may be present, misclassification bias was minimized by examining only major diagnostic categories (broad discharge categories) and total admissions. The SPARCS dataset provides details of utilization and health status by individual, such as diagnosis, as well as other demographic variables such as age, race, and gender for every hospitalization in New York City (and in New York State).  Therefore, while subject to misclassification bias, this dataset is not subject to sampling error. A comprehensive source of ambulatory care utilization data is not available by for New York City.

Mortality analysis

Mortality rates were calculated using data from the 1990 United States Census Public Use Micro-Sample (PUMS) and 1990 Vital Statistics data from the New York City Department of Health.  The error associated with smaller but more recent census samples, such as the Current Population Survey, was prohibitively large for a mortality analysis.

Study Design

In this section, we will discuss the design and limitations of our analysis.  More detailed information will be provided in the “Discussion” section following the presentations of the study results.

Hospital care utilization and morbidity analysis

Country of birth is not routinely recorded in most medical records, including hospitalization data, posing challenges to research on health care utilization and outcomes among the foreign-born.  However, hospitalization records from the SPARCS files do include the address of each patient.  Translating these addresses into the area of residence for each patient, it is possible to analyze community-based trends in hospital use and outcomes for groups of individuals living in specific geographically defined regions of New York City.  Because the foreign-born tend to be concentrated in specific geographic regions of the city, it is possible to analyze differences in hospitalization patterns in neighborhoods with large numbers of foreign-born persons relative to neighborhoods with low numbers of foreign-born persons. Table 1 presents the proportions of the population comprised of immigrants by neighborhood.

We conducted a multivariate linear regression analysis on 55 New York City neighborhoods at the “sub-borough” level.  We first estimated mean hospital admission rates for 55 neighborhoods in New York City by dividing the total number of hospitalizations in a given neighborhood by the number of persons residing in that neighborhood.  Using these neighborhoods as our units of analysis, we estimated correlation coefficients for demographic and socioeconomic covariates with average hospitalization rates by neighborhood.  We then controlled for these factors using multivariate models in order to distinguish the independent effect of the proportionate population level of foreign-born persons residing in the neighborhood on hospitalization rates in that neighborhood.   This research method, termed a “small area analysis,”70,71  allows for the analysis of the independent effect of the percentage of foreign-born persons living in a neighborhood on  hospitalization rates, when individual data are not available, by exploiting the fact that foreign-born persons tend to cluster into ethnic enclaves (demographically similar geographically confined communities that are demographically different from surrounding areas).

All analyses were conducted using the Statistical Package for the Social Sciences (SPSS) version 9.0 for Windows or SPSS version 10 for the Macintosh.  With the exception of total household income, all variables were converted to area-specific proportions by dividing the parameter in question by the total population.  Thus, the proportion of females, foreign-born (total and by sub-group), persons of each race, persons of Hispanic ethnicity, persons living with more than 1.5 occupants per room, households with rats, and persons born in specific region of the world were tabulated for each of the 55 neighborhoods under study and included as independent variables.

The sample size of the Housing and Vacancy Survey (HVS) limited the width of the age-specific intervals we were able to use in our analysis. Moreover, due to sample size limitations, age could not be added as an independent predictor in our linear regression model.  Therefore, four separate models were constructed, each evaluating a different age group.  The age intervals used in the analysis of hospitalization rates were 18 to 25, 25 to 45, 45 to 65 and 65 and older.  We will therefore discuss the results of each of these analyses separately in the “Results” section below.

Because the combined 18 to 45 age group in the HVS data set included large numbers of foreign-born persons who resided in well-demarcated areas, we used this age group in our “base-case” analysis–an analysis that included a descriptive evaluation of foreign-born persons from different regions of the world residing in New York and an examination of major diagnostic categories for hospitalizations.  To ensure that the use of this large age interval did not influence our model, we also analyzed trends for the two smaller (18 to 25 and 25 to 45 intervals).

Approximately 13 percent of the persons sampled in the housing and vacancy survey did not report their country of birth.  We did not include these persons in our calculations.  Instead, we determined the proportion of foreign-born persons living in a neighborhood by dividing the number of persons who self-reported having been born outside of the United States and its territories by the total number of persons (including those that failed to state their country of birth).  The proportion of persons who failed to report their country of origin in a neighborhood was positively associated with income and negatively associated with hospitalization rates.  In other words, healthier persons with high income were less likely to state their country of birth in the housing and vacancy survey.

We tested three age-specific multivariate models of the determinants of hospitalization to allow coefficients to vary by different age groups.  In analyses of sub-groups (e.g. persons from Africa), there were some neighborhoods with too few persons in the sample to obtain a reliable estimate of the proportion of foreign-born.  In these instances, the proportion of persons from the region in question was assigned a missing value and the neighborhood was excluded from the analysis altogether, limiting the statistical power of the model.  In the age-specific models that included all foreign-born persons, sufficient numbers were obtained from all sub-borough areas and none of the neighborhoods were excluded for working age foreign-born persons.  However, analyses of sub-groups were only possible in the base-case model (the combined 18 to 45 year-old-age group) and only univariate analyses were possible for foreign-born persons aged 65 years and over.

An analysis of r values, eigenvalues, condition indexes, and variance proportions indicated that all variables were independent with the exception of parameters specific to overcrowding (i.e. the proportion of households in a neighborhood with more than 1.5 occupants per room and the proportion of households in a neighborhood with rats)–these parameters were dependant on total household income and the proportion females in the neighborhood.  All variables entered were approximately normally distributed (by histogram and p-p plots) and none required transformation.
 
Mortality analysis

Age-specific mortality rates were determined by dividing the number of deaths among foreign-born persons, native-born persons, and Puerto Ricans by their respective numbers of persons residing in New York City using 1990 New York City Department of Health Vital Statistics Data and the 1990 Public Use Micro-Sample (PUMS) from the United States Bureau of the Census.  Age-standardized rates were obtained using the total New York City population as a standard population.

Life expectancy was calculated using standard demographic methods.72,73 We used a large initial age interval because there were small numbers of deaths for persons between the ages of 1 and 25.  Life expectancy was calculated at 1 year of age because foreign-born infants are not tabulated in New York City’s Vital Records.  We assumed that persons over the age of 75 would have a life expectancy of 11.1 years (the life expectancy at 75 in the United States) regardless of the country of origin.74  This assumption likely resulted in a slight under-estimate of the life expectancy of persons born in United States territories and a slight over-estimate of the life-expectancy of foreign-born persons because of the survivor effect, a phenomenon in which groups with relatively higher mortality rates at young ages tend to demonstrate relatively lower mortality rates among the elderly.

Results

Health-Related Demographics

In this section, we will discuss known and potential predictors of morbidity and mortality among native-born, foreign-born, and Puerto-Rican-born persons.  We will also describe how these risk factors might influence the rates of morbidity and mortality among these three populations in New York City.  Because the demographic makeup of New York City changes rapidly from year to year, we will provide some 1999 data for the reader’s interest, however, unless otherwise indicated, all data pertain to the 1996 cohort we utilized in our analysis.

In 1999, approximately forty percent of New York City’s inhabitants were foreign-born.75-77   Foreign-born persons residing in New York City differ demographically from foreign-born persons residing in the United States as a whole.78,79  As in other parts of the country, Asians who immigrated to New York City have higher earnings than native-born persons.  However, persons born in Africa, Europe, and Latin America who reside in New York City report earnings similar to native-born New Yorkers.  These data compare to nationwide data, where African and Asian- born populations tend to have a higher income and have a higher level of education and income than the average U.S. citizen.80-82  Latin America-born persons nationwide, on the other hand, average only eight years of primary schooling and have a mean income just above the poverty line.65-67 Undocumented immigrants, recent immigrants, refugees, and elderly immigrants tend to have lower educational attainment than the average documented immigrant both nationwide and in New York City.65-67

Summary measures of demographic characteristics for New York City residents we calculated using the 1996 Housing and Vacancy Survey (HVS) are presented in Table 2.

The majority of foreign-born persons residing in New York City are male largely due to the predominance of males among Asian-born and African-born populations. Caribbean-born populations, though, are predominantly female (see Figure 1).   Because U.S.-born females have higher than average hospitalization rates than males, but lower than average mortality rates, gender can exert a large influence over morbidity and mortality rates.

While some of the difference in hospitalization rates between genders is due to hospitalization for normal pregnancy, differences remain when admissions specific to childbearing are removed from the analysis.  While pregnancy is by no means a pathological medical condition, it does require the consumption of medical resources and, to the extent that immigrant families have more or fewer children than non-immigrant families, factors into health utilization analyses.
See Figure 1.

Foreign-born persons residing in New York City are remarkably racially diverse (see Figure 2).  For example, a large number of African-born New Yorkers self-identify as white and Korean-born New Yorkers report diverse ancestral backgrounds.  Among persons born in Southeast Asian countries, a large proportion are ethnically Chinese.  The extent to which racial characteristics affect the health status of foreign-born populations is unknown.  However, racial discrimination may influence health outcomes and certain diseases are more common in different races (e.g. cystic fibrosis in whites and sickle cell anemia in blacks).
See Figure 2.

Puerto Ricans differ demographically from both native-born populations and foreign-born populations.  For example, almost 65 percent of Puerto Ricans are female (see Figure 1) and earn less than most other groups residing in New York City and a large percentage self-identify as black or African American (see Figure 2).  Demographically, Puerto Ricans are most similar to Dominicans, a group that is also disproportionately female and has low earnings. Given that Puerto Ricans are demographically similar to Dominicans but need not formally immigrate to permanently reside in the United States, it is tempting to examine these two groups as part of a quasi-experimental case-control study.  However, such an analysis is limited by differences in the rates of drug use between these two groups ([Reference Ted Joyce’s low-birth weight chapter here]).

Univariate Analyses

Hospital care utilization and morbidity analysis

18-45 year olds (base-case analysis)

In New York City neighborhoods, hospitalization rates among 18 to 45 year-olds varied from approximately 5.5 hospitalizations per 100 residents per year (the Upper East Side in Manhattan) to 20 hospitalizations per 100 residents per year (Morrisania in the Bronx) in 1996.  Foreign-born persons were represented in all neighborhoods, comprising between just over 10 percent of the Upper East Side in Manhattan to just under 80 percent of Jackson Heights in Queens.

The mean and standard deviations of selected independent variables in our analysis specific to persons aged 18-45 are presented in Table 3.

We examined univariate associations between the rate of hospitalization in any given neighborhood and various risk factors for hospitalization.  In univariate analyses, total household income, gender, race, ethnicity, country of origin, and overcrowding were all significantly correlated with hospitalization rates (see Table 4).   Total household income, the percentage of persons self-identifying as white, and the percentage of foreign-born residents in a neighborhood were negatively associated with the rate of hospitalization of that neighborhood’s residents.
 
Figure 3 illustrates the relationship between the proportion of foreign-born persons residing in a neighborhood and the rate of hospitalization in that neighborhood.  In this figure, the proportion of foreign-born persons residing in any given neighborhood is represented in the pie chart located in the center of the neighborhood and the hospitalization rate for the neighborhood is indicated by its color (with darker colors representing higher hospitalization rates).  For example, Jackson Heights (see arrow), a neighborhood in which just under 80 percent of all residents are foreign-born has a hospitalization rate of less than 6 percent.
See Figure 3.

Of persons aged 18 to 45 years in New York City, approximately 35% percent were foreign-born in 1996.  Foreign-born persons were disproportionately skewed toward the upper end of this age range; approximately 27% of 18 to 25 year olds were foreign-born and approximately 38 percent of 25 to 45 year olds were foreign-born.  There is linear relationship between the proportion of foreign-born living in a neighborhood and the number of hospitalizations of persons from that neighborhood, with predominantly foreign-born neighborhoods having substantially lower hospitalization rates than predominantly native-born neighborhoods.  Of the 27 neighborhoods with lower than average hospitalization rates, 20 had foreign-born populations above the mean for New York City.  Exceptions included the Upper East Side, the Upper West Side, South Shore, Turtle-Bay, Greenwich Village, Park Slope and Co-op City.  With the exception of South Shore and Co-op City, these neighborhoods are affluent.

Our analyses of foreign-born populations from specific geographic regions were hampered by the size of the Housing and Vacancy Survey sample.  It was only possible to obtain a general idea of which groups have the lowest rates–none of the results were statistically significant.  Neighborhoods represented by persons born in Caribbean nations (other than United States territories) tended to have lower than average hospitalization rates as did neighborhoods with large numbers of persons born in Mexico or the former Soviet Union (two growing foreign-born groups in New York City) and Asia (a large but highly localized community).  Neighborhoods with large populations of persons born in Africa had hospitalization rates similar to those of native-born neighborhoods as did neighborhoods primarily populated by persons born in Western Europe.

Diagnosis-Specific Utilization

Increasing percentages of foreign-born occupants in a neighborhood predicted lower rates of hospitalization for infectious disease, cancer, circulatory conditions, mental illness, and nervous system conditions.  Hospitalization rates for digestive conditions, however, were not significantly lower in neighborhoods with high proportions of foreign-born persons.

Surprisingly, the negative association between the percent of foreign-born persons and hospitalization rates was weaker for cancer than for all diseases but digestive conditions.  With the exception of gastric cancer, the rate of cancer in foreign-born persons is generally reported as lower in the medical literature.29,30,44 Moreover, the rate of most cancers is higher in countries with established market economies than virtually anywhere else in the world.83

It may initially seem surprising that the rate of infectious disease in neighborhoods with large foreign-born populations is lower than in neighborhoods that are predominantly native-born given, given the prevalence of infectious disease in developing nations.83  However, most infectious diseases are of limited duration.  Moreover, legal immigrants are screened for HIV before entering the country, and are excluded from entry if positive.

Between the years 1990 and 1999, 147,513 refugees immigrated to New York City.84  Though refugee populations may have higher rates of mental illness, such as post-traumatic stress disorder and depression, hospitalization rates for mental illness were inversely correlated with the percent of foreign-born persons residing in these neighborhoods.  It is possible that the lower rates of mental illness in other foreign-born groups were sufficiently low to blunt the impact of large refugee populations in some neighborhoods.  It is also possible that the refugees–who are predominantly Jewish persons escaping religious persecution in the former Soviet Union rather than victims of war–have lower rates of mental illness than other refugee groups.

Though foreign-born persons from developing countries may be at greater risk for rare infectious neurological conditions, such as a parasitic infection of the brain called neurocysticercosis, it is not surprising that the overall hospitalization rate for neurological conditions is lower in predominantly foreign-born neighborhoods.  The vast majority of neurological conditions are chronic in nature and severe conditions are likely to be noted during overseas medical examinations used to exclude immigrants attempting to enter the United States legally.  Nor is it surprising that the rate of hospital admissions for heart disease and other circulatory conditions (the most frequent hospital discharge diagnosis) is lower, given that the United States has high prevalence of these conditions relative to other countries.68

There was no association between the percent of foreign-born persons residing in a neighborhood and admissions for digestive disorders.  We were not able to determine the extent of variability among the individual diagnoses.  Some diseases, such as inflammatory bowel disease and diverticulitis, that are common in the United States are extremely rare in developing nations or in the developed nations of Asia.68  On the other hand, cancer of the stomach and parasitic diseases that may sometimes be erroneously classified as digestive rather than infectious, are more common in other countries.

45-65 year olds

Thirty-six percent of New Yorkers aged 45 to 65 years of age were foreign-born in 1996.   Though the total number of subjects in this analysis was considerably smaller than in our analysis of the 18 to 45 age group, the inverse relationship between the percentage of foreign-born persons in a given neighborhood and hospitalization rates remained strong (see Table 5).

There were no overarching differences in hospitalization trends, discussed above for the 18-45-year-old group, for the 45-65 year old group, however, the association between country of birth and hospitalization rate was stronger and the association between income, gender, and race was weaker than for the 18 to 45 year old age cohort.  This trend generally held for the association between the country of birth and diagnosis-specific admissions.

Persons 65 and Over
 
Among persons over the age of 65, approximately 27 percent were foreign-born.  Of these, the vast majority were born in European countries (both Eastern and Western).  With the exception of persons from the former Soviet Union, many of the foreign born in this age group immigrated to the United States in their youth.

Though we found a negative correlation between the percent of foreign-born persons in a neighborhood and hospitalization rates, the association was minimal and not statistically significant.  In inter-group comparisons of health, the gap between the healthiest group and the least healthy group tends to narrow as the population ages.  It is hypothesized that only the healthiest members of disadvantaged groups survive to older ages.  In analyses of mortality data, this phenomenon has been termed the “survivor effect.”  The survivor effect has been observed in analyses applying to survival for persons of differing races, education levels, and income levels.85

Multivariate analyses

Hospitalization among persons aged 18-45

In our multivariate analyses, it was necessary to examine broad age intervals because there were not enough persons aged 18 to 25 in the HVS to obtain reliable estimates of the proportion of foreign-born persons.  Within the 18 to 45 age interval, foreign-born persons tended to be older than native-born persons, suggesting that the inverse relationship between the percentage of foreign-born persons residing in a neighborhood and hospitalization rates is stronger than observed here.  We examined the smaller 25 to 45 age group to determine whether the relationship between the number of foreign-born persons residing in a neighborhood and the hospitalization rate for that neighborhood strengthened.  In this 25 to 45 age group, the proportion of foreign-born persons in any neighborhood explained 12.5 percent  of the variation in hospitalization rates between New York City neighborhoods (relative to 9 percent for the 18 to 45 year old group, see Table 6), and the strength of association increased.  Table 6 presents results of the analysis for the age group 18-45.

In Table 6, the “ß” value indicates the direction and strength of the relationship between the independent variables and proportion of foreign-born persons in a neighborhood, the p-value indicates whether or not the relationship is statistically significant, and the R2 value indicates the proportionate amount of variance explained by the contribution of the independent parameter in explaining neighborhood hospitalization rates.

Native-born New Yorkers aged 18-45 are more likely to be hospitalized than foreign-born persons after controlling for covariates and excluding Puerto Ricans from the analysis.  When income, gender, race, and ethnicity are not accounted for, the risk of hospitalization for native born persons, while higher than average, was not statistically significant in our analysis.

Controlling for income, the negative association between the proportion of foreign-born persons and hospitalization rates becomes stronger.  Given this, it is likely that the earnings of the foreign-born in New York City has an impact on their health, as it does for native-born Americans.  After controlling for income, the percent of foreign-born persons residing in a neighborhood explains approximately 22 percent of the variation in hospitalization rates between neighborhoods.

After controlling for income, ethnicity, race, and gender, higher percentages of foreign-born persons residing in a neighborhood predict even fewer hospitalizations.  Most foreign-born groups are either predominantly white or Asian.  Among the foreign-born who self-report as black, race appears to play a small role as a determinant of a neighborhood’s hospitalization rate (as in Africans) or predicts fewer hospitalizations (as in Caribbeans).

Foreign-born males, like native-born males, are at lower risk of hospitalization than females, even when normal pregnancies are not included in the analysis (data not shown).  However, gender appears to be a weaker predictor of hospitalization among the foreign-born than it is among native-born persons living in New York City.  A good part of this difference appears to be explained by demographics; native-born females tend to live in overcrowded households, have a low household income, are more likely to report the presence of rats in their households, and are more likely to be Hispanic or black than native-born males.  When these factors are controlled for, native-born females and foreign-born females have similar hospitalization rates.

Neighborhoods with large Asia-born populations–a group with a higher income than native-born New Yorkers–did not in of itself explain lower hospitalization rates in neighborhoods with high numbers of foreign-born persons.  Most neighborhoods with high proportions of Asian immigrants had similar hospitalization rates to predominantly foreign-born neighborhoods with lower proportions of Asia-born persons.  Moreover, controlling for Asian ethnicity had little impact on the association between the total number of foreign-born persons residing in a neighborhood and the hospitalization rate.

The factors we studied explained approximately 69 percent of the variation in the rate of hospitalization between neighborhoods, with income accounting for over 54 percent of the hospitalization rate and the percentage of foreign-born persons explaining almost 10 percent of the variation.  Though females are significantly more likely than males to be hospitalized (due in part to childbearing and in part to higher overall rates of illness), gender composition explained only three percent of the association between hospitalization rates and the characteristics of a neighborhood.

Immigrants who naturalize often reside in established immigrant neighborhoods and it was not possible to control for the effect of lower hospitalization rates in established immigrant neighborhoods.  Therefore, estimates of hospitalization rates in neighborhoods with high proportions of recently naturalized neighborhoods are imperfect.  Nonetheless, it is likely that recently naturalized citizens are at least as healthy as other foreign-born persons residing in New York City, if not healthier.

Recently arrived persons have a lower mean household income than established immigrants but may have not been exposed to potentially harmful environmental factors such as crime or a high fat diet.  To analyze the impact of recent immigration (regardless of citizenship status) on hospitalization rates, we examined 1999 data from the Housing and Vacancy Survey, which included questions pertaining to the person’s length of residence in the United States, and compared this to 1996 hospitalization data.

The proportion of foreign-born persons residing in a neighborhood who arrived after 1990 was associated with a similar decrement in hospitalization rates as the proportion of persons who arrived before 1990, even when income, gender, and race are not controlled for in the analysis.
 

Hospitalization among persons aged 45 to 65
 
The direction, magnitude and significance of the determinants of hospitalization for persons aged 45-65 years old were highly similar to those for the 18-45 year old group. When controlling for gender, income, race and income, the percent of foreign-born persons aged 45 to 65 was negatively correlated with the rate of hospitalization (ß = -0.082; p < 0.0001).   In our analysis of this age group, we were able to account for approximately 61 percent of the factors leading to hospitalization.

As in the analysis of persons aged 18 to 45, the percentage of Puerto Ricans in a neighborhood predicted significantly higher hospitalization rates (ß = 0.13; p = 0.001), as did the percentage of persons born in the continental United States (ß = 0.067; p = 0.003).

As in the 18 to 44 group, it was not possible to conduct an analysis of neighborhoods with large numbers of persons born in specific geographic regions.

Hospitalizations among persons aged 65 and over

As noted in the section presenting the results of our univariate analyses above, there were no significant differences in hospitalization rates between neighborhoods when only persons over the age of 65 were examined.  This held true when we attempted to control for race, income, and gender, however, the statistical power of these analyses was inadequate to predict significant differences between neighborhoods.  When only persons over the age of 65 were examined in the model, race, gender, and income also proved to be poor independent predictors of hospitalization.  We hypothesize that only weak differences were observed due to the survivor effect.

Costs

Using hospital charges specific to each age group and the linear regression equation for the percent foreign-born, it is possible to estimate differences in hospitalization costs associated with foreign-born hospitalization patterns.  In 1996, approximately 35 percent of New York’s inhabitants were born in a foreign country.  That year, hospitalization charges for persons 18 to 45 were approximately 3.5 billion dollars.  If the foreign-born had hospitalization patterns similar to native-born populations, an additional $327 million dollars in hospital charges would have been incurred in 1996 (95 percent confidence interval $318 million to $327 million).
 
Hospital charges are not an accurate indicator of actual societal costs of medical care but, to the extent these charges are reimbursed, they do reflect the burden of medical care on insurance policy holders and taxpayers.  Using cost to charge ratios, which are derived using data from the Health Care Financing Administration, it is possible to estimate the burden of this care on society as a whole.  After adjusting for actual societal costs, the hospitalization patterns among foreign-born persons in this age group resulted in costs that were $136 million dollars lower than hospital costs would have been for a similar group of native-born New York City residents.

The differential in hospital charges between native-born and foreign-born hospitalization patterns of persons in the 45 to 65 age group was greater than in the 18-45 age group, amounting to $423 million (95 percent confidence interval $287 million to $535 million) on total charges of 3.2 billion dollars.  It is possible that the larger differential in the 45 to 65 age group despite similar overall charges reflects an underestimation of “savings” in the younger groups since foreign-born persons in that group were older than native born persons. The overall “savings” realized by society, due to the lower hospitalization rates among the foreign born, were approximately 176 million (95 percent confidence interval $119 million to $223 million).  It was not possible to estimate the costs or savings associated with immigration for the 65 years and over group.
 
Mortality Rates

While the foreign-born appear to have lower rates of hospitalization than native-born persons, it is possible that the foreign-born are less likely to seek care or be referred to a hospital when ill.  If so, all-cause mortality rates should be higher for foreign-born persons than for native-born persons.  If the foreign-born are healthier than native-born persons, it is likely that mortality rates would also be lower.

We examined all-cause age-specific mortality rates using 1990 census data and 1990 mortality data from the New York City Department of Health.  We also calculated age-adjusted mortality rates and life expectancy for foreign-born persons, native-born persons, and persons born in territories of the United States.

Foreign-born persons have lower age-standardized rates of death, a longer life expectancy, and fewer years of life lost to disease than native-born persons (see Table 7).  In 1990, foreign-born persons would be expected to live four years longer than native-born persons at one year of age and a full six years longer at one year of age.  This, coupled with the lower hospitalization rates we observed earlier, suggests that the foreign-born are healthier than native-born persons.

The mortality rate for foreign-born persons between the ages of 1 and 25 years of age was not significantly different from rates reported for native-born persons, however.  Since we were not able to include younger persons in our hospitalization study, it is not possible to examine the relationship between hospital utilization and mortality for this group.

Persons born in territories of the United States residing in New York City appear to be at greater risk of death than persons born in the United States when all age groups are considered together.  However, after the age of 45, the risk of death is roughly comparable to persons born in the continental United States and Hawaii.

Discussion

Given the lower hospitalization and mortality rates among foreign-born persons living in New York, the foreign-born appear to be in better health than native-born New Yorkers.  Hospitalization rates for cancer, circulatory disorders, mental illness, neurological conditions, and infectious disease all decline as the percentage of foreign-born persons in a neighborhood increases and this trend continues until at least age 65.  Though real or perceived barriers may partially explain lower hospitalization rates to medical care, mortality rates are also generally lower for foreign-born persons than native-born persons in New York City.

The lower hospitalization and mortality rates among foreign-born groups may in part be due to selection factors among immigrant groups, since legal immigrants must undergo a medical examination (to rule out excludable conditions, mostly active infectious diseases) prior to entry into the United States and only healthy persons are able to travel.  It may also be due to lower rates of chronic disease in the countries from which the people originate.  Given that most chronic diseases emerge later in life, are difficult to detect on the overseas medical examination, and account for the majority of hospitalization costs, the latter likely explains the bulk of the lower rate of illness.  On the other hand, it is likely that the affluence of the foreign-born in New York City (relative to other persons in the countries from which they immigrate), travel requirements, and the overseas medical examination, explains the lower rates of hospitalizations for infectious disease among immigrant groups relative to native-born groups.  Moreover, persons with chronic disease may be more likely to develop bacterial pneumonia (the most prevalent infectious disease in the United States) than healthy persons.

Three distinct factors may contribute to the higher rates of mortality and comparable rates of morbidity in foreign-born persons over the age of 65 we observed.  First, foreign-born persons may experience a decline in health status as they become acculturated and are exposed to risk factors for chronic disease, such as a poor diet, poor housing conditions, and crime.  In support of this hypothesis, we found that the neighborhoods with larger proportions of immigrants who naturalized (became citizens of the United States) in the year 1996 had rates of hospitalization that were comparable to or lower than established foreign-born neighborhoods. Second, there are clear demographic differences in the various age cohorts we studied with the younger groups mostly having arrived from developing countries and the older groups predominantly having arrived from Western Europe and, more recently, Western Europe.  Finally, foreign-born persons tend to survive to older ages than native-born and Puerto-Rico-born persons resulting in a de facto selection bias for the healthiest persons from the latter two groups.  This “survival effect” is frequently seen in longitudinal cohort studies.86

There was no association between the proportion of foreign-born or the proportion of native-born persons residing in a neighborhood and the proportion of persons who failed to disclose their country of birth.  Both the HVS and the 1990 census likely under-estimate the number of foreign-born persons (especially undocumented persons), impoverished persons, and racial minority groups.  We assumed that this underestimate was consistent across sub-borough areas.  If this assumption is correct, undercounting would have no effect on the small area analysis since only the proportion of foreign-born persons residing in each sub-borough area was entered into the multiple regression models.

Given that the foreign-born are likely to be undercounted in the United States Census,87 it is possible that the mortality rates among the foreign-born we observed represent a high estimate since population data is used as the denominator of such rates..  It is not likely that undercounting affected our analysis of hospitalization rates, since that analysis relied upon percentages rather than total numbers.  However, to the extent that the foreign-born are less likely to be counted in poorer neighborhoods than in more affluent neighborhoods, it is possible that these figures, too, represent an underestimate.

We conducted a sub-analysis of citizenship effects using 1990 census data to determine whether undercounts of undocumented immigrants could have skewed the results.  In this analysis, citizenship was positively correlated with hospitalization rates at the census tract level, suggesting that undocumented immigrants may be at lower risk than naturalized persons.  If this is the case, our analysis may further underestimate the negative association between hospitalization rates and the proportion of foreign-born persons in a neighborhood.

In the 18 to 45 age group, we likely underestimated the overall decrement in hospitalization rates, since foreign-born persons tended to be older than native-born persons within the interval.  This may also account for the stronger negative association between hospitalization rates in foreign-born neighborhoods and the larger savings in hospital charges we observed in the 45 to 65 age group.

It is possible that the finding of lower hospitalization rates among the foreign-born merely represents reduced access to needed medical services and the lower mortality rates reflect emigration of very ill persons to their country of birth.  The hospitalization rates we measured reflect the total number of hospitalizations occurring among residents of a given neighborhood in a single year divided by the total population of that neighborhood.  If severely ill persons return to their homeland, this might also lower the overall rate of hospitalization in predominantly foreign-born neighborhoods.  However, the diagnosis of a grave illness necessitates diagnostic medical visits, suggesting that, in these instances, such persons would be accessing the medical system.  Moreover, it is likely that people with severe medical conditions would prefer to receive care in the United States rather than in less technologically advanced countries. To date, we are not aware of any studies examining whether emigration for severe illness is a common phenomenon.

Though we were not able to estimate hospitalization costs or savings associated with foreign-born populations for persons under the age of 18 or over the age of 65, it is not likely that hospital costs associated with either age group differ greatly from native-born persons since mortality rates in both of these age groups were similar to native-born persons.  If this is the case, savings in hospital charges alone amounted to almost three-quarters of a billion dollars in New York City in 1996 alone.  The overall societal savings, which reflect costs associated with the production and delivery of products and services but not costs to taxpayers or to government institutions, was over 313 million dollars.  Were we able to include the cost of ambulatory and long-term care, the savings may have been substantially greater.

It is possible that the demographic composition of some foreign-born enclaves could explain the lower overall hospitalization rates in predominantly foreign-born neighborhoods.  Persons arriving from Asia, for example, tend to have above-average incomes, and tend to be predominantly male and the foreign-born in general are more likely to self-report being white–factors associated with lower hospitalization rates.  On the other hand, foreign-born persons living in New York City are also somewhat more likely than average to be Hispanic (a description of ethnicity rather than race), to live in overcrowded situations, and to be of lower than average income.  Controlling for all of these variables together strengthens the association between increased proportions of foreign-born living in a neighborhood and reduced hospitalization rates suggesting that, on the whole, the demographic profile of foreign-born populations places such populations at higher risk of hospitalization.

Recent research suggests that income distribution may be a stronger predictor of health outcomes than total household income.  It is hypothesized that an individual’s relative position in society is a stressor that results in subsequent health effects. This may explain why relatively egalitarian societies, such as the state of Kerala in India, tend to be world leaders in life expectancy despite abject poverty.88,89  It may also explain why impoverished communities that are geographically proximate to affluent communities, such as Harlem, tend to rank among the worst in the world despite relative affluence (Harlem is in the top 20th percentile worldwide in terms of purchasing power and total household income).90  With respect to income distribution, foreign-born persons tend to fare better than native-born persons (see Figure 4).

With the exception of persons from the Dominican Republic and the former Soviet Union, New York City’s foreign-born have more favorable income distribution than native-born New Yorkers.  Foreign-born Asian groups enjoy a markedly egalitarian pattern of income distribution. While twenty-two percent of native-born New Yorkers earned under $15,000 in 1996, only four percent of Filipino-born persons earned under that amount.

Moreover, only 18 percent of Filipino-born New Yorkers earned under $35,000 that year.
New Yorkers born in Puerto Rico (a United States territory) and the Dominican Republic have substantially lower earnings than native-born New Yorkers.  Persons from other Caribbean nations, who are largely of African descent (see Figure 2), generally have higher total household earnings and have a more favorable distribution of income than their Dominican and Puerto Rican neighbors.  It is not possible to make accurate comparisons between persons born in the former Soviet Union and native-born New Yorkers because they generally immigrate as refugees and are eligible for financial support from the United States’ government.

It has been observed that, as foreign-born communities age into first and second generation enclaves, income distribution patterns begin to mirror those of the surrounding native-born communities and health indicators, such as infant mortality and life expectancy, tend to appear the same as, or worse than, nearby predominantly native-born communities.91  No causal pathway has been established for the link between the decay in income distribution patterns and health indicators, however, and it is quite possible that such associations are spurious.

Our findings clearly cannot be understood without considering the well-being of foreign-born communities.  The dramatic imbalances in gender composition coupled with relatively large numbers of married foreign-born persons in of New York City suggests that many persons residing in the city have left loved ones in the countries they left, for example.  Moreover, the high rates of overcrowding observed in predominantly foreign-born communities suggests that housing conditions are less than optimal.  Our study results must be viewed with the consideration that health is better defined as a state of well-being rather than the presence or absence of disease.

Moreover, lower rates of hospitalization among the foreign-born in no way suggests that there is less a need for access among immigrant populations; despite their apparent reduced risk of disease, foreign-born persons still become ill and still require access to needed medical services. The pattern of reduced rates of illness among foreign-born populations merely suggests that the cost of such programs would be lower, rather than higher, than services currently granted to native-born persons.

Conclusion

In conclusion, there is no evidence to suggest that foreign-born residents in New York City have lower levels of health status nor higher levels of hospitalization than do native-born New York City residents.  The hypothesis that the foreign-born are “draining” U.S. resources, due to disproportionately high rates of disease, is rejected in this analysis.

Health care policy in New York City is at a pivotal point concerning how best to address the complex relationship between health, health care spending, and economic productivity among vulnerable immigrant groups and minorities. Traditional resource allocations among health programs may need to be re-examined as immigrants and their families develop predictable health needs over the life cycle.  It is important to recognize the dynamic nature of health status among immigrants.  The degree of risk upon arrival, which  depends on the health risk profile of the country of origin, will shift to new health risks post arrival, including, for instance, emotional stresses that impact on mental health, environmental concerns such as lead poisoning as a consequence of immigrants living in substandard housing, or occupational hazards.  In addition, the children and families of recent immigrants pose additional disease prevention challenges.

New models of care may need to be developed to improve access and utilization for those immigrants who do have medical service needs, and cost-effective alternative ways of providing culturally appropriate preventive screening and health education programs may need to be implemented.  Policy makers should address these concerns without hesitation.  Appropriately meeting the health needs of immigrants will not disproportionately reduce the availability of health resources for New York City’s native-born population.