Abstract
Background:
Using healthcare, whether for routine preventative examinations, chronic condition management, or emergent conditions, is an essential element of achieving and maintaining health. Over 44 million migrants live in the US today and nearly half (44.6 %) are Latino. To the extent that immigration law-related concerns deter US Latino migrants from using healthcare, they jeopardize the health of a substantial though marginalized US subpopulation.
Methods:
A multistate sample (N = 1750) of noncitizen adult, Spanish speaking Latino migrants (both undocumented and documented) living in the US completed a cross-sectional survey comprising a scale assessing perceptions of immigration laws and consequences related to healthcare use and whether they had received and/or needed but did not receive healthcare in the previous 12-months. Participants were recruited in community settings and by word-of-mouth. Univariate analyses examined associations among study outcomes and common predictors of healthcare use. Multivariable analyses examined the relative contribution of perceived immigration laws and consequences on healthcare use and unmet need.
Results:
Perceptions of immigration laws and immigration consequences were a significant predictor of not having received healthcare in the previous 12-months and having needed and not received healthcare in the same period, even when considered relative to common predictors of healthcare use. Immigration documentation status and preferred language predicted healthcare use in univariate analyses but not in the multivariable model.
Conclusions:
Perceived immigration laws and consequences related to healthcare use may influence migrants’ healthcare use. Effective interventions should be developed to address immigration-related concerns.
Keywords: Healthcare use, Latino migrants, Immigration policy, Immigration concerns
1. Introduction
Over 44 million migrants live in the US today, nearly half of whom (44.6 %) are Latino [1]. The majority of US migrants are native to Latin American countries and especially Mexico [1]. There is widespread concern about migrants’ financial burden on the US healthcare system [2-5]. Yet in general, US Latino migrants use fewer healthcare resources and contribute more resources to US healthcare systems than the resources they use [4-6]. Noncitizen adult Latino migrants, and migrants in general, visit doctors less frequently, are less likely to have a regular healthcare provider, and receive fewer preventative healthcare services [7] than both US-born Latinos and non-Latino whites [6-9].
Yet healthcare use is essential to achieving the self-reliance and quality of life that is central to US immigration policy and to migrants’ own goals [10]. Underutilization of care leads to reduced opportunities to prevent illness, detect emergent conditions before they become dire, and manage chronic conditions [11].
Researchers and commentators have identified several common drivers of the underutilization of healthcare among Latino migrants. These include large numbers of noncitizen migrants lacking health insurance, inadequate care or lack of providers for Spanish-speaking persons with limited English proficiency, and employment in low-level and often unregulated jobs where health benefits are not available [12-14]. Researchers have also identified the influence of generalized immigration fears on deterred healthcare use [12,15].
However, we know little about the specific nature of migrants’ immigration concerns related to healthcare use, the accuracy of these concerns, and importantly, the relative contribution of immigration fears on the underutilization of care given general common, migrant-specific and general contributors to underutilization of care.
2. Methods
Results reported here were derived from a multistate study (N = 1750) on the influence of actual and perceived immigration-related laws on Latino migrants’ willingness to use, and actual use of, services for HIV testing, and for two additional drivers of HIV infection–substance use disorders, and intimate partner violence. Healthcare use in the previous year and unmet need for healthcare in the same timeframe were also assessed. The current analysis addresses this healthcare use and unmet need. The IRB of the first author’s institution approved all research procedures.
2.1. Sample
Participants were adult, sexually active, Spanish-speaking, non-citizen Latino migrants (documented and undocumented) who had been living in the US for at least 12 months. (N = 1750). Being sexually active was required to assess primary study outcomes reported elsewhere. To examine participants’ healthcare use in the US in the previous year, only those who had lived in the US for one year or more were included in the present analysis.
2.2. Procedures
Cross-sectional data were collected between 9/29/19 and 1/19/21 from Latino migrants living in four US cities—two cities in self-identified immigration sanctuary states, Chicago, IL, Los Angeles, CA, and two cities in states that did not identify as sanctuary locations, Phoenix, AZ, and Raleigh, NC. These cities were selected because they represented diverse immigration environments.
Participants were recruited through word-of-mouth and on-line distribution of the study flyer. Our goal was to recruit a heterogeneous sample of Latino migrants that resembled the respective city’s Latino migrant population in terms of age, immigration status, self-identified gender, and country of birth. As a result, we distributed flyers in community sectors where diverse subgroups of Latino migrants lived and worked and among community-based organizations that served diverse Latino migrant sub-populations.
Promotional flyers encouraged Latino migrants to call or text a number indicated on the flyer to be screened for eligibility. All screening occurred over the telephone in conversations with research team members. Flyers were disseminated widely online (e.g., through Facebook posts) and by community-based organizations serving migrants. Flyers were also posted in places likely to be seen by Latino migrants such as ethnic markets and restaurants. Project staff also conducted outreach in community settings (e.g., recreation centers, places where day laborers congregate, consulates, and local fast-food restaurants in areas with substantial Latino migrant residents). In this case, participants were screened for eligibility on site and if eligible, completed the survey at that time.
From 9/23/19 to 3/15/20, the survey was self-administered in-person via ACASI on electronic tablets provided by outreach workers. From 7/16/20 to 1/19/21, after the COVID-19 pandemic emerged, the survey was self-administered through the Qualtrics survey platform on individual’s own internet-connected devices. Eligible participants were sent a survey link and an informational letter via cellular phone text or email. In cases where on-line self-administration was a barrier to participation, interviewers administered the survey via telephone. For in-person survey administration, eligible participants were scheduled after screening to complete the survey at a time and location convenient for them or, as mentioned, completed the survey at community sites directly after screening.
To assuage concerns about confidentiality, participants’ names were not collected during screening or survey administration, whether in-person or on-line. For on-line surveys, participants were told that they could further protect their anonymity by creating an email account specifically for the study. IP addresses were deleted from the survey database. The survey took approximately 60–90 minutes to complete. Eligible, interested persons gave informed consent at screening after a review of the study informational letter with a research team member. After completing the survey, participants watched a debriefing video addressing common legal misconceptions and were given information on relevant local resources. Participants then received a $50 incentive in the form of cash or gift card. All materials were in Spanish. All procedures were approved by the Internal Review Board at the first author’s institution.
2.3. Measures
2.3.1. Healthcare use
Participants’ healthcare use was assessed by a single item: “Have you received any medical care in the US in the previous 12 months?” (‘yes’; ‘no’). An additional item assessed whether participants had had an unmet medical need in the last 12-months: “In the last 12 months, was there ever a time when you needed medical care but did not receive it?” and if so, reasons why they did not receive it ranging from “providers in my area do not speak Spanish” to “clinics in my area do not accept my health insurance.” Visits to hospital emergency departments (EDs) for care were assessed with the item “How many times did you go to an emergency department or the emergency room at a hospital in the last 12 months? (dichotomized for analysis as once or more and never). The primary reason for the ED visit was also elicited “Which best describes why you went to the emergency department or emergency room at a hospital?” Responses ranged from “I had an accident or sudden illness” to “I can only go to the doctor after work hours”.
2.3.2. General sociodemographic variables
Sociodemographic variables assessed included participant’s age, self-identified sex, education, income, the metropolitan area where they lived, and whether they had health insurance. (For analysis, metropolitan area was dichotomized into cities within states self-identified as immigration sanctuaries (Chicago IL and Los Angeles CA) and those within states that were not immigration sanctuaries (Phoenix AZ and Raleigh NC) [16].
2.3.3. Self-rated health and access to health information
Self-rated health and access to health information were assessed with single items. Self-rated health was assessed by the item “In general, how would you rate your overall physical health?” Response options were on a 4-point scale: ‘very good’, ‘good’, ‘average’, and ‘poor’. Responses were dichotomized for analysis: ‘very good’/‘good’ vs. ‘average/‘poor’. Access to information on health was assessed with a single item: “I know where to get my questions about health answered”. Response options were dichotomous (‘yes’; ‘no’).
2.3.4. Immigration-specific sociodemographic variables
These variables included country of origin, (dichotomized for analysis as born in Mexico; not born in Mexico), language preference (dichotomized to speaks English always or most of the time; other than English always or most of the time), immigration documentation status (dichotomized for analysis as having any documentation; not having any documentation). Participants also indicated the number of years they had resided in the US (dichotomized to 1–5 years; >5 years).
2.3.5. Immigration enforcement concerns and deportation experience
Immigration enforcement concerns was assessed with an adapted version of Arbona et al.‘s (2010) acculturative stress measure [18]. The scale asks about activities avoided because of an individual’s fears of deportation. These ranged from “Do you avoid asking for help from government agencies because you have concerns about being deported?” and “Do you avoid reporting a crime that someone has done to you to the police because you have concerns about being deported?” Response options were dichotomous (‘yes’; ‘no’) (range 0–7; Cronbach’s alpha (α) = 0.84) Deportation experience was assessed with four items adapted from the Hispanic Stress Inventory [17]. The items asked whether participants had been questioned by immigration authorities, detained by immigration authorities, had been deported, had a close family member or friend be deported, or feared deportation. Response options were dichotomous (‘yes’; ‘no’).
2.3.6. Perceived local immigration environment, perceived discrimination, and access to immigration-related information
Perceived local immigration climate was assessed with seven items. These ranged from “Immigration policies in my city make immigrants feel unwelcome” to “Immigration authorities in my city are trying to make all immigrants leave the state”. Response options were on a four-point Likert-type scale ‘disagree’, ‘somewhat disagree’, ‘somewhat agree’, and ‘agree’, (range 7–28; α = 0.78). Perceived discrimination was assessed with a 9-item ‘Everyday Discrimination Scale’ [19]. Participants indicated the frequency with which they experienced discrimination in their daily lives. Items ranged from being treated with less courtesy than others to being threatened or harassed. Response options were ‘never’, ‘less than once a year,’ ‘a few times a year,’ ‘a few times a month,’ ‘at least once a week,’ and ‘every day’, (range 9–54; α = 0.93).
Two questions assessed participants’ access to immigration-specific information. One item assessed participants’ access to information on immigration law, “I know where to get information about immigration law,” and one on access to information on public services available to immigrants, “I know where to get information about services available to immigrants.” Response options were dichotomous (‘yes’; ‘no’).
2.3.7. Perceptions of immigration laws and immigration consequences related to healthcare use
Participants’ perceptions of immigration laws relevant to health care use and the immigration ramifications of using healthcare were assessed with a 16-item de novo scale (range 16–64; α = 0.88) Items ranged from “If an immigrant uses services at a community health center, they can seem as if they cannot support themselves and hurt their immigration status” to “Doctors and other health service workers are required to report undocumented immigrants to immigration officials.” All items were erroneous, reflecting common misperceptions of immigration-related laws. Response options were on a 4-point scale: ‘disagree’, ‘somewhat disagree’, ‘somewhat agree’, and ‘agree’.
2.4. Analyses
2.4.1. Descriptive statistics
Descriptive statistics (frequency and percent or mean and standard deviation) were calculated to characterize sample socio-demographics, healthcare use, need, and history, self-rated health, perceived information access, perceived immigration climate, deportation experience, fear of deportation, perceived discrimination, and perceived immigration laws and consequences related to general health.
2.4.2. Missing data for items and scales
When a participant’s response was missing for a scale item, the sample median of the non-missing responses for each item was imputed. If a participant left more than 15% of the items within a scale missing, the computed scale score was assigned missing. Cronbach’s alpha was computed for each scale using the current study data.
2.4.3. Multivariable analysis
Logistic regression analyses were conducted to examine factors associated with not having received healthcare in the previous 12-months and having needed care and not received it in the previous 12-months. Factors with a Wald χ2 p-value of <.20 in univariate regression were included in multivariable regression. Non-significant factors with the largest Wald p-value ≥.05 were removed one-by-one until each remaining factor had a p-value <.05 to determine independent factors associated with healthcare use. The odds ratio and 95 % confidence interval (CI) associated with a one-unit increase in the value of a predictor is reported. Analyses were conducted using IBM SPSS Statistics, Version 26.
3. Results
Table 1 displays sample characteristics divided into general characteristics (i.e., those not specific to migrants) and migrant-specific factors. The mean age of participants was 37.9 years (SD 10.6, range 18–77). Most identified as female (64.1 %, n = 1122) and most did not have health insurance (79.6 %, n = 1393). Participants were evenly sampled from among the four metropolitan areas. Over half reported that they were undocumented (58.2 %, 1019) and three-quarters spoke Spanish or another non-English language all or most of the time (75.1 %, n = 1313).
Table 1.
Sociodemographic characteristics of participants.
| Variable | Labels | N (%)a |
|---|---|---|
| Metropolitan area | Los Angeles CA | 466 (26.6) |
| Phoenix AZ | 431 (24.6) | |
| Chicago IL | 428 (24.5) | |
| Raleigh NC | 425 (24.3) | |
| Age, years | Mean (SD) | 37.9 (10.6) |
| Missing | 0 (0.0) | |
| Self-identified gender | Male | 597 (34.1) |
| Female | 1122 (64.1) | |
| Transgender | 10 (0.6) | |
| Other | 2 (0.1) | |
| Missing | 19 (1.1) | |
| Education completed | Less than 6th grade | 270 (15.4) |
| 6th grade but not completed high school | 566 (32.3) | |
| Graduated high school/received GED | 462 (26.4) | |
| Graduated technical degree | 128 (7.3) | |
| Attended some college | 143 (8.2) | |
| College graduate | 137 (7.8) | |
| Missing | 44 (2.5) | |
| Currently has health insurance | Yes | 299 (17.1) |
| No | 1393 (79.6) | |
| Missing | 58 (3.3) | |
| Monthly income | $0-$499 | 376 (21.5) |
| $500-$999 | 387 (22.1) | |
| $1000-$1999 | 552 (31.5) | |
| $2000-$4999 | 298 (17.0) | |
| $5000 or more | 20 (1.1) | |
| Missing | 117 (6.7) | |
| Physical health self-rating | Very good | 369 (21.1) |
| Good | 777 (44.4) | |
| Average | 490 (28.0) | |
| Poor | 78 (4.5) | |
| Missing | 36 (2.1) | |
| I know where to get my questions about health answered. | Yes | 1263 (72.2) |
| No | 425 (24.3) | |
| Missing | 62 (3.5) | |
| Received any medical care in the US in the previous 12 months | Yes | 829 (47.4) |
| No | 873 (49.9) | |
| Missing | 48 (2.7) | |
| Ever a time needed care and didn’t receive it in the previous 12-months | Yes | 239 (13.7) |
| No | 1457 (83.3) | |
| Missing | 54 (3.1) | |
| Used emergency department or emergency room at a hospital in last 12 months, among those who received any medical care (n = 829) | Yes | 288 (34.7) |
| No | 537 (64.8) | |
| Missing | 4 (0.5) | |
| Times went emergency department or emergency room at a hospital, last 12 months, among those who received any medical care (n = 829) | 537 (64.8) | |
| Once | 193 (23.3) | |
| More than once | 95 (11.4) | |
| Missing | 4 (0.5) | |
| Preferred language | English always or most of the time | 65 (3.7) |
| English and Spanish about the same | 344 (19.7) | |
| Spanish always or most of the time | 1305 (74.6) | |
| Some other language always or most of the time | 8 (0.5) | |
| Missing | 28 (1.6) | |
| Country of origin | México | 1297 (74.1) |
| Guatemala | 101 (5.8) | |
| El Salvador | 98 (5.6) | |
| Honduras | 87 (5.0) | |
| Dominican Republic | 34 (1.9) | |
| Cuba | 32 (1.8) | |
| Ecuador | 23 (1.3) | |
| Nicaragua | 6 (0.3) | |
| Another country | 46 (2.6) | |
| Missing | 26 (1.5) | |
| Documentation status | Permanent or other documented status | 526 (30.1) |
| Undocumented immigrant | 1019 (58.2) | |
| Missing | 205 (11.7) | |
| Immigration recency | Over 5 years | 1143 (82.5) |
| 5 years or fewer | 302 (17.3) | |
| Missing | 5 (0.3) | |
| Perceived local immigration climate, range 7–28 (high = less welcoming) | Mean (SD) | 16.5 (5.7) |
| Missing | 122 (7.0) | |
| Any deportation experience | Yes | 824 (47.1) |
| No | 825 (47.1) | |
| Missing | 101 (5.8) | |
| I know where to get information about immigration law. | Yes | 963 (55.0) |
| No | 716 (40.9) | |
| Missing | 71 (4.1) | |
| I know where to get information about public services available to immigrants. | Yes | 915 (52.3) |
| No | 776 (44.3) | |
| Missing | 59 (3.4) | |
| Immigration enforcement concerns scale, score range 0–7; higher scores represent greater concerns | Mean (SD) | 2.0 (2.2) |
| Missing | 110 (6.3) | |
| Discrimination scale, score range 9–54; higher scores represent greater perceived discrimination | Mean (SD) | 18.7 (10.4) |
| Missing | 107 (6.1) | |
| Perceived immigration laws and consequences, score range 16–64; higher scores represent greater perceived immigration consequences from healthcare use | Mean (SD) | 38.2 (11.2) |
| Missing | 141 (8.1) |
N = 1750.
The majority of participants rated their health as very good or good (66.5 %, n = 1146). Half of participants reported not receiving healthcare in the previous 12-months (49.9 %, n = 873). A total of 239 participants (13.7 %) reported that there was at least one time in the previous 12-months when they needed care but did not receive it. The most common reasons for not receiving care when it was needed were “I didn’t have health insurance” (n = 77) and “I didn’t know where to go” (n = 50). To explore whether migrants appeared to use hospital emergency departments (EDs) as a primary source of care, data were collected on use of EDs. Of those receiving healthcare in the previous year, just over one-third received care in an ED at least once (34.7 %, n = 288). Most of the persons who used the ED used it only once (67 %, n = 193). The primary reason given for using an ED was “I had an accident or sudden illness” (n = 185).
Participants’ immigration enforcement concerns were modest with a mean of rating of 2.0 (s d. 2.2) on a scale of 0–7 with higher scores representing more acute enforcement concerns. Still, nearly half had some deportation experience (48.2 %, n = 821). Overall, participants neither wholly agreed or disagreed that their local immigration climate did or did not welcome migrants. On a scale of 7–28 with higher scores reflecting perceptions of a less welcoming environment, the mean score was 16.5 (s d. 5.7).
Results revealed widespread misunderstanding of immigration laws and consequences. A total of 1647 (94.1 %) of participants agreed or somewhat agreed with one or more erroneous statements about restrictions on immigrants’ healthcare use and legal consequences arising from using healthcare. Still, responses to the scale as a whole tended toward “somewhat disagree”. With a possible range of scores from 16 to 64 (higher scores indicating perceptions of more restrictive laws and deleterious consequences), the mean score was 38.2 (s d.11.2) (coefficient of variation 0.29).
Table 2 reports univariate and multivariable predictors of not having received healthcare in the previous 12-months. In univariate analysis, general sociodemographic predictors of not having received healthcare in the previous year were consistent with findings of previous studies. Those not receiving healthcare were younger (odds ratio (OR 1.03, 95 % confidence interval (CI) (1.02–1.04) p < .001), identified as male (OR 2.22 (1.80–2.73) p < .001), had less than a high school education (OR 1.44 (1.19–1.74) p < .001), and reported not having health insurance (OR 2.88 (2.20–2.77) p < .001). Those living in Phoenix or Raleigh-Durham were more likely to not have received care (OR 1.72 (1.40–2.08) p < .001). Those rating their health status as ‘average’ or ‘poor’ were less likely to not have received care (OR 0.66 (0.54–0.81) p < .001). Immigration-related factors significantly predicting not having received healthcare included being an undocumented migrant (OR 1.26 (1.02–1.56) p = .030), speaking Spanish or other non-English language most or half of the time (OR 2.11 (1.25–3.56) p = .005), and being a recent immigrant (OR 1.92 (1.48–2.49) p < .001). Those having deportation experience were less likely to report not having received care (OR 0.70 (0.57–0.84) p < 001). Greater scores on the de novo perceived immigration law and consequences scale, indicating more misperceptions of immigration-related laws and more negative perceptions of laws and anticipated consequences, was a significant predictor not having used healthcare (OR 1.02 (1.01–1.03) p < .001).
Table 2.
Univariate and multivariable associations with did not receive any medical care in the U.S. in the past 12 months.
| Univariatea | Multivariablea,c | |||
|---|---|---|---|---|
| OR (95 % CI) |
p | aORa (95 % CI) |
padj | |
| General factors | ||||
| Study location | <.001 | <.001 | ||
| Los Angeles/Chicago | 1.0 | 1.0 | ||
| Phoenix/Raleigh-Durham | 1.72 (1.40–2.08) | 1.57 (1.25–1.98) | ||
| Age, years | 0.97 (0.96–0.98) | <.001 | 0.98 (0.97–0.99) | .002 |
| Self-identified genderb | <.001 | <.001 | ||
| Female | 1.0 | 1.0 | ||
| Male | 2.22 (1.80–2.73) | 1.98 (1.56–2.51) | ||
| Education | <.001 | .005 | ||
| Less than HS education completed | 1.44 (1.19–1.74) | 1.39 (1.11–1.75) | ||
| Graduated from high school/received a GED or beyond | 1.0 | 1.0 | ||
| Health insurance | <.001 | <.001 | ||
| Currently has health insurance | 1.0 | 1.0 | ||
| No health insurance | 2.88 (2.20–3.77) | 3.07 (2.26–4.18) | ||
| Monthly income | .84 | |||
| $0-$499 | 1.0 | – | ||
| $500-$999 | 1.18 (0.89–1.58) | – | ||
| $1000-$1999 | 1.12 (0.86–1.45) | – | ||
| $2000-$4999 | 1.08 (0.79–1.46) | – | ||
| $5000 or more | 1.02 (0.42–2.51) | – | ||
| Overall physical health rating | <.001 | <.001 | ||
| Very good/Good | 1.0 | 1.0 | ||
| Average/Poor | 0.66 (0.54–0.81) | 0.65 (0.51–0.82) | ||
| I know where to get my questions on health answered. | <.001 | <.001 | ||
| Yes | 0.27 (0.21–0.35) | 0.35 (0.27–0.46) | ||
| No | 1.0 | 1.0 | ||
| Immigration-related factors | ||||
| Preferred language | .005 | – | ||
| Speaks Spanish or other non-English language most or half of the time | 2.11 (1.25–3.56) | – | ||
| Speaks English always or most of the time | 1.0 | – | ||
| Undocumented immigrant | .030 | – | ||
| Yes | 1.26 (1.02–1.56) | – | ||
| No | 1.0 | – | ||
| Country of origin | .007 | |||
| Born in Mexico | 0.74 (0.59–0.92) | – | ||
| Born outside the US, other than in Mexico | 1.0 | – | ||
| Immigrated to the US within the previous 5 years | <.001 | – | ||
| Yes | 1.92 (1.48–2.49) | – | ||
| No | 1.0 | – | ||
| Perceived local immigration climate scale, range 7–28 (high = less welcoming) | 1.03 (1.02–1.05) | <.001 | – | |
| Deportation experience | <.001 | .001 | ||
| Yes | 0.70 (0.57–0.84) | 0.67 (0.54–0.84) | ||
| No | 1.0 | 1.0 | ||
| I know where to get information on immigration law. | <.001 | – | ||
| Yes | 0.52 (0.43–0.64) | – | ||
| No | 1.0 | – | ||
| I know where to get information on public services available to immigrants. | <.001 | – | ||
| Yes | 0.54 (0.44–0.66) | – | ||
| No | 1.0 | – | ||
| Discrimination scale, score range 9–54; higher scores represent greater perceived discrimination | 1.04 (1.03–1.05) | <.001 | – | |
| Immigration enforcement concerns scale, score range 0–7; higher scores represent greater concerns | 1.05 (1.00–1.09) | .045 | – | |
| Perceived immigration laws and consequences, score range 16–64; higher scores represent greater perceived immigration consequences from healthcare use | 1.02 (1.01–1.03) | <.001 | 1.02 (1.01–1.03) | <.001 |
Regression Model Chi-square 286.0, df 9, p < .001; Nagelkerke R Square 0.23; Hosmer Lemeshow Goodness of Fit Chi-square 8.32, df 8, p = .40; With AUC optimal cut point 0.4787, overall percentage correct 67.3, sensitivity 67.9 % and specificity 68.1 %; Positive predictive value 68.9 %.
Adjusted Odds ratios (aOR) are adjusted for other variables in the multivariable model.
N = 1702; Excludes 48 individuals who did not answer if they had received medical care in the U.S. in the past 12 months.
Excludes 10 transgender and 2 with other gender identity.
In multivariable analysis of not having received healthcare in the previous 12-months, the general sociodemographic factors in univariate analysis continued to be significant. Of immigration-related factors, only two variables emerged as significant predictors of not having received healthcare in the previous 12-months: not having deportation experience (OR 1.49 (1.18–1.87) p = .001) and greater scores on the perceived immigration laws and consequences scale (OR 1.02 (1.01–1.03) p=<.001).
Table 3 displays univariate and multivariable analyses of needing healthcare but not receiving it in the previous 12-months. In univariate analyses, the only significant general sociodemographic predictor of not receiving care when needed was living in Phoenix or Raleigh-Durham (OR 1.32 (1.01–1.74) p = .044). Being female tended toward significance (OR 1.35 (0.99–1.82) p = .053). Persons who rated their health as ‘average’ or ‘poor’ as opposed to ‘very good’ or ‘good’ were also more likely to have needed but not received care. (OR 2.35 (1.78–3.10) p < .001). Surprisingly, not having health insurance did not predict unmet healthcare need. On the other hand, most of the immigration-related variables were significant in univariate analysis. Notably, preferring Spanish or another language over English and being an undocumented migrant were not among these significant predictors.
Table 3.
Univariate and multivariable associations with In the last 12 months, was there ever a time when you needed medical care but did not receive it.
| Univariatea | Multivariablea,c | |||
|---|---|---|---|---|
| OR (95 % CI) |
p | aORa (95 % CI) |
padj | |
| General factors | ||||
| Study location | .044 | – | ||
| Los Angeles/Chicago | 1.0 | – | ||
| Phoenix/Raleigh-Durham | 1.32 (1.01–1.74) | – | ||
| Age, years | 1.01 (0.99–1.02) | .19 | – | |
| Self-identified genderb | .053 | |||
| Female | 1.35 (1.00–1.82) | – | ||
| Male | 1.0 | – | ||
| Education | .22 | |||
| Less than HS education completed | 0.84 (0.64–1.11) | – | ||
| Graduated from high school/received a GED or beyond | 1.0 | – | ||
| Health insurance | .086 | – | ||
| Currently has health insurance | 1.0 | – | ||
| No health insurance | 1.41 (0.95–2.08) | – | ||
| Monthly income | .26 | |||
| $0-$499 | 1.0 | – | ||
| $500-$999 | 0.65 (0.43–1.00) | – | ||
| $1000-$1999 | 0.96 (0.67–1.39) | – | ||
| $2000-$4999 | 0.87 (0.56–1.34) | – | ||
| $5000 or more | 1.32 (0.43–4.10) | – | ||
| Overall physical health rating | <.001 | <.001 | ||
| Very good/Good | 1.0 | 1.0 | ||
| Average/Poor | 2.35 (1.78–3.10) | 1.98 (1.47–2.68) | ||
| I know where to get my questions on health answered | .79 | – | ||
| Yes | 1.04 (0.76–1.44) | – | ||
| No | 1.0 | – | ||
| Immigration-related factors | ||||
| Preferred language | .95 | – | ||
| Speaks Spanish or other non-English language most or half of the time | 1.02 (0.50–2.10) | – | ||
| Speaks English always or most of the time | 1.0 | – | ||
| Undocumented immigrant | .35 | – | ||
| Yes | 1.16 (0.85–1.57) | – | ||
| No | 1.0 | – | ||
| Country of origin | .65 | |||
| Born in Mexico | 0.93 (0.68–1.27) | – | ||
| Born outside the US, other than in Mexico | 1.0 | – | ||
| Immigrated to the US within the previous 5 years | .14 | – | ||
| Yes | 0.75 (0.51–1.10) | – | ||
| No | 1.0 | – | ||
| Perceived local immigration climate scale, range 7–28 (high = less welcoming) | 1.09 (1.06–1.12) | <.001 | – | |
| Deportation experience | <.001 | <.001 | ||
| Yes | 2.50 (1.86–3.36) | 1.85 (1.34–2.54) | ||
| No | 1.0 | 1.0 | ||
| I know where to get information on immigration law | .003 | – | ||
| Yes | 0.66 (0.50–0.87) | – | ||
| No | 1.0 | – | ||
| I know where to get information on public services available to immigrants | 0.049 | – | ||
| Yes | 0.76 (0.58–1.00) | – | ||
| No | 1.0 | – | ||
| Discrimination scale, score range 9–54; higher scores represent greater perceived discrimination | 1.03 (1.01–1.04) | <.001 | – | |
| Immigration enforcement concerns scale, range 0–7 (high = greater enforcement concerns) | 1.24 (1.17–1.32) | <.001 | 1.15 (1.08–1.23) | <.001 |
| Perceived immigration laws and consequences, score range 16–64; higher scores represent greater perceived immigration consequences from healthcare use | 1.04 (1.03–1.06) | <.001 | 1.03 (1.01–1.04) | <.001 |
Regression Model Chi-square 106.31, df 4, p < .001; Nagelkerke R Square 0.12; Hosmer and Lemeshow Goodness of Fit Test Chi-square 9.91, df 8, p = .27; With optimal cut point 0.134, overall percentage correct 64.2 %, sensitivity 64.2 % and specificity 64.2 %; Positive predictive value 22.8 %.
Adjusted Odds ratios (aOR) are adjusted for other variables.
N = 1696; Excludes 54 individuals who did not answer if “In the last 12 months, was there ever a time when you needed medical care but did not receive it?”
Excludes 10 transgender and 2 with other gender identity.
In multivariable analysis, only one general demographic predictor was significantly associated with not having received care when it was needed: rating one’s health as ‘average’ or ‘poor’ vs ‘good’ or ‘very good’. (OR 1.98 (1.47–2.68) p < .001). Variables relevant to migrants again emerged as significant predictors including having greater immigration enforcement concerns (OR 1.15 (1.08–1.23) p < .001) and having had deportation experience (OR 1.85 (1.34–2.54) p < .001). Once again, greater scores on the perceived immigration law and consequences scale (indicating greater agreement with erroneous statements about immigration laws and greater perceived consequences of using healthcare) predicted needing but not having received care in the previous 12-months (OR 1.03 (1.01–1.04) p < .001). Absent were undocumented status and preference for Spanish. Having health insurance was also not a predictor.
4. Discussion
In our multivariable analysis of health care use and gaps in care, common, nonimmigration-related factors associated with increased need for care emerged such as age and poor self-rated health status. Since those who need care are more likely to use it, and to have instances when they need but do not receive care, this finding is not surprising.
Similarly, our findings indicated that those who had deportation experience were more likely to have received healthcare in the previous 12-months and more likely to have needed care but not receive it. The deleterious effect of deportation on health has been noted in other research [20,21]. It is likely that persons who have deportation experience have more healthcare needs resulting in receiving care more often, and because of this increased need, needing care but not receiving it as well.
Negative perceived immigration laws and consequences related to healthcare emerged as a significant predictor of both not utilizing health services and needing but not receiving care. This association emerged even after considering common, nonimmigration-related contributors to healthcare use. Indeed, when considering those who needed but did not receive care in the previous year, the only general sociodemographic factor emerging as significant was age.
Further, perceptions of immigration laws and consequences related to healthcare remained a significant predictor of underutilization of care and unmet healthcare needs even when considering common immigration-related barriers to care including undocumented status and limited resources for non-English speakers. These multivariable analysis results are contrary to expectations and indicate a need for further analysis for confirmation, especially when misperceptions of laws are considered. Nearly half of our participants (829; 47.4 %) reported receiving healthcare in the US in the previous year. Given that many of these participants were undocumented, uninsured, and had lower incomes, sources of care for many particularly medically marginalized persons appear to have been available, at least in some instances. Our analysis suggests that unmet healthcare needs may be determined by migrants’ concerns about accessing resources that are available as well as by lack of access to resources based on demographic characteristics. Importantly, that resources for care may be available does not mean that the care received by marginalized persons is optimal or timely. In-depth study on quality of care among persons who are undocumented, uninsured, and have lower incomes is desperately needed.
The data on which this analysis was based were collected during a particularly tumultuous time in US immigration policy, especially related to the care of undocumented migrants and those who were recent US permanent residents. In Winter of 2019, a draft amendment to US immigration guidance, the public charge rule, was circulated. In Fall of 2020, the amendment was adopted (84 FR 41292). The topics were therefore of particular salience to participants. In Spring of 2021, the 2019 rule was vacated (86 FR 14221). Within the patchwork of laws and rules that constitute US immigration policy and the political rhetoric that surrounds the topic, it is likely that migrants form perceptions of consequences related to healthcare use based on broad public communications and anti-immigrant sentiment. Unfortunately, immigration policy in the US has not stabilized since data were collected, suggesting that any deleterious consequences of this political context likely persist.
The misperceptions of immigration-related laws revealed in this study underscores migrants’ need for accurate information on relevant laws. Given that erroneous beliefs about laws and consequences may be more amenable to immediate intervention through education than larger and often immutable factors examined, this finding could be encouraging from an intervention perspective. Education from a trusted source may increase healthcare use in the immediate future while larger, more pernicious structural issues are addressed.
Two additional immigration-related predictors were significantly associated with needing but not receiving healthcare in the previous year–perceived discrimination and local immigration enforcement concerns. Those with unmet needs appear to be among the most alienated migrants in the sample and among those perceiving the greatest immigration risk. Unstable laws, contentious legal battles, and widely disseminated anti-immigrant political rhetoric may well contribute to a sense of alienation among immigrants with the consequence of deterring healthcare use.
5. Limitations
Data were derived from a convenience sample and were collected by self-report. These introduce the possibility of bias. Still 58.2 % of the sample were undocumented, pointing to diversity with a critical factor in studies of migrants. Our sample comprised more women than men. Given that women receive more healthcare than men, results may be biased toward women’s perspectives on healthcare use.
Conclusion
In this large, multi-state sample of US Latino migrants, inaccurate perceptions of immigration laws and consequences of using healthcare predicted not having used healthcare and not receiving healthcare when needing it, even given other common predictors of healthcare utilization. This demonstrates the importance of this often-overlooked influence on Latino migrants’ healthcare use. More research on the influence of deportation experience and health is needed. On the other hand, those with deportation experience were more likely to have received care in the previous year and also more likely to report needing care but not receiving it. This also calls for further research into the effect of deportation on health. Further, research is also needed to confirm findings that predictors commonly found to influence healthcare use such as undocumented status and lack of health insurance are less of an influence when misperceptions of laws and need for information on available resources are considered as our findings suggest.
Acknowledgements
We would like to acknowledge the contribution of the following Proyecto Luz 2.0 research team members: Ruzanna S. Aleksanyan, MS, Julia Lechuga, PhD, Nora Bouacha, MPP, Andrea L. Dakin, PhD, Joanna L. Barreras, PhD, MSW, Samantha Garza, MS, Sara LeGrand, PhD, Celina Lopez, Elizabeth Ortiz de Valdez, MD, Juan Reyes, Silvia Valadez-Tapia, MA, Angel Rosado, and Juan Flores.
Statements and declarations
Research reported in this document was supported by the National Institute on Minority Health and Health Disparities under award number R01MD011573. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
CRediT authorship contribution statement
Carol L. Galletly: JD PhD was involved in all aspects of the manuscript including conceptualization, Funding acquisition, Project administration, Supervision, Formal analysis, Methodology, Validation, Writing – original draft, and, Writing – review & editing. Timothy L. McAuliffe: PhD was involved in the conceptualization, Formal analysis, Methodology, Validation, Writing – original draft, and, Writing – review & editing. Julia B. Dickson-Gomez: PhD was involved in the, Conceptualization, of the manuscript and, Writing – review & editing. Laura R. Glasman: PhD was involved in the, Conceptualization, of the manuscript and, Writing – review & editing. Dulce M. Ruelas: DrPH was involved in the, Conceptualization, of the manuscript and, Writing – review & editing.
Declaration of competing interest
No potential competing interest was reported by the authors.
References
- [1].Esterline C, Batalova J. Spotlight frequently requested statistics on immigrants and immigration in the United States. https://www.migrationpolicy.org/article/frequently-requested-statistics-immigrants-and-immigration-united-states; March 14, 2023. 3.2023.
- [2].Ku L. Assessing the presidential proclamation on visas and health insurance. Health Affairs Blog; 2019; Dec 17. 10.1377/hblog20191217.16090. . [Accessed 12 September 2022]. [DOI] [Google Scholar]
- [3].Felter C, Renwick D, Cheatham A. The US immigration debate. Updated June 23, 2020, https://www.cfr.org/backgrounder/us-immigration-debate-0. [Accessed 20 May 2020].
- [4].Muschek AG. The impact of undocumented immigrants on the healthcare infrastructure of the United States. J Soc Polit Econ Stud 2015;40(4):329–68. winter. [Google Scholar]
- [5].Flavin L, Zallman L, McCormick D, Wesley Boyd J. Medical expenditures on and by immigrant populations in the United States: a systematic review. Int J Health Serv 2018. Oct;48(4):601–21. 10.1177/0020731418791963. [DOI] [PubMed] [Google Scholar]
- [6].López-Cevallos D. Are Latino immigrants a burden to safety net services in nontraditional immigrant states? Lessons from Oregon. Am J Publ Health 2014. May;104(5):781–6. 10.2105/AJPH.2013.301862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Himmelstein J, Himmelstein DU, Woolhandler S, Bor DH, Gaffney A, Zallman L, Dickman S, McCormick D. Health care spending and use among hispanic adults with and without limited English proficiency. Health Aff 2021. Jul;40(7):1126–34. 10.1377/hlthaff.2020.02510. 1999-2018. [DOI] [PubMed] [Google Scholar]
- [8].Ortega AN, McKenna RM, Kemmick Pintor J, Langellier BA, Roby DH, Pourat N, Vargas Bustamante A, Wallace SP. Health care access and physical and behavioral health among undocumented Latinos in California. Med Care 2018. Nov;56(11):919–26. 10.1097/MLR.0000000000000985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Derose KP, Bahney BW, Lurie N, Escarce JJ. Review: immigrants and health care access, quality, and cost. Med Care Res Rev 2009. Aug;66(4):355–408. 10.1177/1077558708330425. [DOI] [PubMed] [Google Scholar]
- [10].Galletly CL, Barreras JL, Lechuga J, Glasman LR, Cruz G, Dickson-Gomez JB, Brooks RA, Ruelas DM, Stringfield B, Espinoza-Madrigal I. US public charge policy and Latinx immigrants’ thoughts about health and healthcare utilization. Ethn Health 2023. Jan;28(1):96–113. 10.1080/13557858.2022.2027879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].US Department of Health and Human Services (HHS). Healthy people 2030, https://health.gov/our-work/national-health-initiatives/healthy-people/healthy-people-2030.last. March/2023.
- [12].Pérez-Escamilla R, Garcia J, Song D. Health care access among Hispanic immigrants:¿ Alguien está escuchando?[Is anybody listening?]. NAPA Bull 2010. Nov 1;34(1):47–67. 10.1111/j.1556-4797.2010.01051.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Rangel Gómez MG, López Jaramillo AM, Svarch A, Tonda J, Lara J, Anderson EJ, Rosales C. Together for health: an initiative to access health services for the hispanic/Mexican population living in the United States. Front Public Health 2019. Sep 24;7:273. 10.3389/fpubh.2019.00273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Velasco-Mondragon E, Jimenez A, Palladino-Davis AG, Davis D, Escamilla-Cejudo JA. Hispanic health in the USA: a scoping review of the literature. Publ Health Rev 2016. Dec;37(31). 10.1186/s40985-016-0043-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Martinez O, Wu E, Sandfort T, Dodge B, Carballo-Dieguez A, Pinto R, Rhodes SD, Moya E, Chavez-Baray S. Evaluating the impact of immigration policies on health status among undocumented immigrants: a systematic review. J Immigr Minority Health 2015. Jun;17(3):947–70. 10.1007/s10903-013-9968-4. Erratum in: J Immigr Minor Health. 2016 Feb;18(1):288. Rhodes, Scott [corrected to Rhodes, Scott D]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Vaughan JM, Griffith B. Map: sanctuary cities, counties, and states. Center for Immigration Studies; 2023. Jul. https://cis.org/Map-Sanctuary-Cities-Counties-and-States. [Google Scholar]
- [17].Cervantes R, Padilla A, Snyder NS. The hispanic stress inventory: a culturally relevant approach to psychosocial assessment. Psychol Assess 1991. Sep;3(3):438–47. 10.1037//1040-3590.3.3.438. [DOI] [Google Scholar]
- [18].Arbona C, Olvera N, Rodriguez N, Hagan J, Linares A, Wiesner M. Acculturative stress among documented and undocumented Latino immigrants in the United States. Hisp J Behav Sci 2010. Aug;32(3):362–84. 10.1177/0739986310373210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Williams DR, Yu Yan, Jackson JS, Anderson NB. Racial differences in physical and mental health: socio-economic status, stress and discrimination. J Health Psychol 1997. Jul;2(3):335–51. 10.1177/135910539700200305. [DOI] [PubMed] [Google Scholar]
- [20].Hamilton ER, Orraca-Romano PP, Vargas VE. Legal status, deportation, and the health of returned migrants from the USA to Mexico. Popul Res Pol Rev 2023, March;42(16). 10.1007/s11113-023-09745-7. [DOI] [Google Scholar]
- [21].Morris JE, Palazuelos D. The health implications of deportation policy. J Health Care Poor Underserved 2015. May;26(2):406–9. 10.1353/hpu.2015.0038. [DOI] [PubMed] [Google Scholar]