The adverse health effects of punitive immigrant policies in the United States: A systematic review

Background Immigrants in the United States (US) today are facing a dynamic policy landscape. The Trump administration has threatened or curtailed access to basic services for 10.5 million undocumented immigrants currently in the US. We sought to examine the historical effects that punitive laws have had on health outcomes in US immigrant communities. Methods In this systematic review, we searched the following databases from inception–May 2020 for original research articles with no language restrictions: Ovid MEDLINE, Ovid EMBASE, Cochrane Library (Wiley), Web of Science Core Collection (Clarivate), CINAHL (EBSCO), and Social Work Abstracts (Ovid). This study is registered with PROSPERO, CRD42019138817. Articles with cohort sizes >10 that directly evaluated the health-related effects of a punitive immigrant law or policy within the US were included. Findings 6,357 studies were screened for eligibility. Of these, 32 studies were selected for inclusion and qualitatively synthesized based upon four themes that appeared throughout our analysis: (1) impact on healthcare utilization, (2) impact on women’s and children’s health, (3) impact on mental health services, and (4) impact on public health. The impact of each law, policy, mandate, and directive since 1990 is briefly discussed, as are the limitations and risk of bias of each study. Interpretation Many punitive immigrant policies have decreased immigrant access to and utilization of basic healthcare services, while instilling fear, confusion, and anxiety in these communities. The federal government should preserve and expand access for undocumented individuals without threat of deportation to improve health outcomes for US citizens and noncitizens.

Introduction Immigration is an irrefutable and foundational tenet of the American nation. The United States (US) continues to maintain considerable immigrant populations, with more than 44.7 million immigrants residing within the country in 2018 [1]. Contemporarily, immigrants in the US are experiencing a rapidly changing policy landscape. The Trump administration's restrictive policies including its "zero tolerance" immigration policy, the narrowing of asylum qualifications, and expansions of the "public charge" definition to include any individual who uses or is likely to use one public benefit, have dramatically narrowed and delayed the pathway to citizenship or permanent resident status and thus access to basic services for the estimated 10.5 million undocumented immigrants currently in the US [2].
Most recently, the astounding racial and ethnic disparities emerging between the victims of the 2020 SARS-CoV-2 pandemic are notable. Preliminary analyses from New York City have shown that Latinx individuals have the highest mortality (22.8%) of any reported ethnic group [3]. In fact, in 20 of 45 US states reporting ethnic data of COVID-19 cases, the proportion of cases among Latinx individuals is at least twice as high as what would be expected on the basis of population; in 11 of the 45 states with available data, the proportion of Latinx individuals infected is more than three times as high [4,5]. It is important to note, that these data do not distinguish between native, naturalized, noncitizen, or undocumented Latinx individuals; as can be expected, data on the morbidity and mortality of undocumented individuals are nearly impossible to gather. However, given that 44% of US immigrants report having Hispanic or Latinx origins, it can be taken as a certainty that the current pandemic has greatly and disproportionally affected immigrant communities [1]. Previously, the definition of "public charge" included Supplemental Security Income, Temporary Assistance for Needy Families, state general assistance programs, and long-term, publicly funded institutionalization. The current administration has added non-emergency federally funded Medicaid, the Supplemental Nutrition Program known colloquially as "food stamps," Section 8 housing assistance, public housing, and state and local cash assistance. Of note, while the Trump Administration had suspended enforcement of the new public charge rule on July 29, 2020, over four months after the onset of the COVID-19 pandemic in the US, the law was reinstated as enforceable on September 11, 2020 [6]. As cases of COVID-19 continue to hit record highs across the US, the Trump Administration has yet to resuspend enforcement of this rule [7]. As such, immigrants in the US today are faced with the unprecedented disadvantage of having to navigate a global pandemic in the setting of an exceedingly hostile immigration climate.
To examine the historical effects anti-immigrant policies have had on health outcomes for communities they directly affect, we systematically selected and reviewed studies published after the Immigration Act of 1990 that investigated the adverse health effects that specific antiimmigrant policies had on immigrant communities in the US.

Methods
This study was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [8]. In adherence to these guidelines, a protocol was registered in PROSPERO (registration #CRD42019138817).

Search strategy
A medical librarian performed comprehensive searches to identify studies that evaluated local, state, or federal anti-immigrant laws and examined their effects on the health of immigrant communities in the US. We define anti-immigrant law as any law or policy at the local, state, or federal level that serves to restrict immigrant access to basic services, public benefits, or employment, or increase the threat of legal consequence or deportation. Studies that exclusively assessed the effects of positive or inclusive immigrant laws were excluded from analysis, as an analysis of the protective effects of positive immigrant laws was beyond the scope of our research question. However, studies that assessed both positive and punitive laws were eligible for inclusion. Searches were initially run on July 8, 2019, and updated on May 4 2020, in the following databases: Ovid MEDLINE 1 (ALL 1946 to present); Ovid EMBASE (1974 to present); Cochrane Library (Wiley); Web of Science Core Collection (Clarivate Analytics); CINAHL (EBSCO); and Social Work Abstracts (Ovid). Search terms included all subject headings and associated keywords for the concepts of immigrants/migrants and law/policy in the US. Specific punitive legislation was also searched by name. The full search strategy for Ovid MED-LINE is available in Supplemental File 2. No language or article type restrictions were imposed.

Study selection
After de-duplication, two independent reviewers screened 6,357 citations using Covidence systematic review software. Titles and abstracts were reviewed by two authors against pre-defined inclusion/exclusion criteria (Cohen's kappa = 0.324). Studies that evaluated the health-related effects of a specific punitive immigrant policy in the US published after the year 1992 were considered for inclusion. Excluded articles were those that: (1) focused exclusively on the effects of positive, inclusive, or protective immigrant laws; (2) were outside the US; (3) had a small patient cohort (n<10); (4) evaluated a potential, rather than demonstrated, health impact; (5) did not study the effects of specific immigrant legislation or policy enactment; and (6) were published before 1992. Discrepancies were resolved by consensus. Due to lack of data to demonstrate impact, conference abstracts, posters, and presentations were also excluded. Full text publications were then obtained for 172 selected studies for a second round of eligibility screening (Cohen's kappa = 0.533). Reference lists of included articles and articles citing included studies were pulled from Scopus (Elsevier) and also screened. The PRISMA flow diagram is available in Fig 1. Data were extracted by three reviewers independently with standardized forms in order to collect the following variables: study setting; population and participant demographics and baseline characteristics; details of the legislation; impact of legislation; follow-up time; follow-up data; study methodology. Due to data heterogeneity, no meta-analysis was performed. Study limitations were reported and risk of bias was assessed and noted.

Results
A total of 32 articles were systematically selected for inclusion in this review (Table 1). These studies were qualitatively synthesized based upon four themes: (1) impact on healthcare utilization, (2) impact on women's and children's health, (3) impact on mental health services, and (4) impact on public health. See Table 2 for a summary of each policy referenced herein.

Impact on healthcare utilization
The 1994 passage of California's Proposition 187 resulted in a significant decrease in new walk-ins (p<0.001) and no-show and patient cancelation rates (p = 0.01) in a study of ophthalmology clinic utilization, both of which only returned to baseline once Proposition 187 was

PLOS ONE
The adverse health effects of punitive immigrant policies in the United States

PLOS ONE
The adverse health effects of punitive immigrant policies in the United States prohibited undocumented individuals from enrolling in any public postsecondary schooling or receiving financial aid; invalidated contracts between parties if one party is knowingly undocumented; established a Class C felony for an undocumented individual entering into a "business transaction" with a government agency; prohibited undocumented individuals from applying for or soliciting work; criminalized "harboring" or "transporting" undocumented individuals; created a state immigration police force (Continued ) stayed [9]. Additionally, while 65% of primary care directors perceived a decline in patient visits or were aware of a person who delayed care due to Proposition 187, their study did not find a significant decrease in primary care clinic visits in its wake [10].
In 1996, both the Personal Responsibility and Work Opportunity Reconciliation Act ("Welfare Act") as well as the Illegal Immigration Reform and Immigrant Responsibility Act (IIR-IRA) were passed; an interview series assessing their effects found that undocumented individuals consistently indicated they would refrain from or delay seeking medical treatment due to fears of immigration enforcement [11]. Another interview series of undocumented immigrants assessed the effects of Proposition 187 and the Welfare Act in both California and Texas, finding that 39% of respondents expressed fear of not receiving medical services due to immigration status. Interestingly, there was no significant decrease in those expressing fear in Texas with respect to California, which may suggest that Proposition 187's effects extended beyond California's border [12].
A 2010 study examining the effects of the Deficit Reduction Act on the utilization of Medicaid family planning services reported a 33% decrease in individuals seeking care through the Family Planning Expansion Project, versus 10% for non-Family Planning Project individuals (p<0.01), with a 47% decline in visits for clients aged <18, compared to a 31% decline for those aged >18 (p<0.01) [13].
A 2013 study assessing the effects of the E-Verify mandate on both voluntary returnees and deportees surveyed at the San Diego-Tijuana border crossing concluded that while E-Verify mandates did not have a significant effect on individuals' ability to obtain healthcare services or government assistance, they did raise deportation fears and reduce interstate mobility among voluntary returnees [14].
A 2017 study investigating undocumented immigrants' and healthcare providers' perceptions of Georgia's HB87, Section 287(g) of the Immigration and Nationality Act (INA), and the Secure Communities Program found that strict anti-immigrant policies increased fear and trauma among immigrants, which resulted in their refraining from or delaying care, restricting mobility, including for therapy or exercise, and seeking alternate care at small, private, fee-forservice clinics [15].
A 2017 Boston interview series of immigrants, healthcare professionals, and immigrant and health advocacy organization employees examining the relationship between municipal health reform and immigrant healthcare access revealed a consensus that immigrants' documentation status hinders their access to healthcare regardless of health coverage status. Thus, even in states with inclusionary policies, federal exclusionary policies and national anti-immigrant sentiment can pose barriers to healthcare [16]. A study investigating how Latinx individuals living in mixed-status households navigate healthcare after SB1070's passage found that over 50% of interviewees reported experiencing difficulty obtaining health coverage and endorsed the complexity of application as a major obstacle. Other barriers to care included discrimination and fear related to deportation and/or detainment while seeking public services, leading to severe delays or complete avoidance of care [17]. Similar concerns were raised in response to Prince William County, Virginia's "Rule of Law" ordinance, with many Latinx individuals expressing concerns of high costs, language difference, and perceived indifference or hostility on the part of healthcare personnel, all of which suggest that these individuals are more inclined to forgo or delay seeking care [18].

Impact on women's and children's health
Three studies assessed the effect of Proposition 187 on Latinx women and children by using proxy markers for undocumented status such as foreign-born status, low educational attainment (<12 years), lack of private insurance, and residing in a county with a high proportion of undocumented immigrants [19][20][21]. In this population, the authors found a significant but small decline in the use of prenatal care (p<0.001), delayed commencement of prenatal care (p<0.001), and a lower number of visits (p<0.001) with respect to those of US-born women of similar educational attainment, with no significant difference in birth outcomes. These women expressed a sense of being "trapped" by their inability to seek publicly funded services, unjustly considered public charges, subject to increased racism and discrimination, and potentially subject to deportation. Combined with the effects of anti-immigrant legislation on the labor market, these women expressed an inability to enter the private market, while now ineligible for public assistance. Interestingly, after the staying of Proposition 187, autism diagnoses for Latinx-born children increased by 25-30%, suggesting an increase in healthcare utilization, which contrasts dramatically with the 13% decrease in autism diagnoses observed during the Proposition 187 period [19][20][21]. Similar concerns were reported by Latinx women in Birmingham in the wake of Alabama's HB56, with many expressing confusion over healthcare eligibility, delays in seeking care, increased use in home remedies, fear of deportation, concerns over costs, fear of driving, and incidences of discrimination and mistreatment by healthcare staff [22].
Two studies examined the effects of the Welfare Act and IIRIRA on immigrant gynecologic and prenatal care. Interviews with safety-net providers, immigrant advocacy organizations, and government agencies revealed a consensus that patients were afraid to apply for public benefits and unable to utilize services in the absence of care, and credit fears of being labeled a "public charge," confusion about the law implementation process, and eligibility requirements as exacerbating factors. While no significant difference was found in the attainment or quality of care when assessing patterns via residency status, the fear burden was significantly higher in undocumented individuals, who were more likely to report having heard rumors that healthcare was unavailable due to immigration status [23,24]. Although the Welfare Act had no direct effect on Medicaid eligibility for low-socioeconomic status children of non-permanent residents, it caused a significant "chilling effect," with Medicaid coverage decreasing by 17% and levels of uninsured children rising by 10% with respect to those of permanent residents between 1996 and 2001 [25]. A similar effect was noted after Georgia's HB87, with a significant decrease in Latinx pediatric patients presenting to the Emergency Department in the post-HB87 period (18.3% vs 17.1%, p<.01), an increase in the percentage of high acuity Latinx patients (14.3% to 16.3%), and a significant increase in Latinx patients admitted post-HB87 (10.2% vs. 8.7%, p<.01). Of note, Latinx was the only ethnic group to see a decrease in visits and increase in acuity in the study [26].
Two studies assessed the effects of SB1070 on Latinx mothers. The law's passage was associated with declines in adolescent use of public assistance (b = -0.51;OR = 0.60; 95% CI = 0.39, 0.92) who were also were less likely to take their baby to the doctor (b = -1.41;OR = 0.24; 95% CI = 0.08, 0.70); compared with older adolescents, younger adolescents were less likely to use preventive healthcare after SB1070. Mother figures were also less likely to use public assistance after SB1070 if they were born in the US and if their post-SB1070 interview was closer to the law's enactment (b = -0.47; OR = 0.63;95%CI = 0.39, 0.99). With regards to birth weight, a significant decline by 15 g (0.52 oz) was observed between July-December 2010, suggesting that these effects were linked to SB1070's signing into law rather than implementation. No significant decreases in birth weight were observed in US-born Latinx, black, or white women [27,28].
While a 2015 study examining how implementation of the Secure Communities Program and section 287(g) of the INA affected prenatal care utilization among Latinx women in North Carolina did not find significant differences in care utilization before and after policy implementation, Latinx mothers were found to delay and receive inadequate care compared to non-Latinx mothers. Additionally, Latinx individuals reported intense mistrust and avoidance of healthcare services [29]. Potochnick and colleagues also looked at the effects of local-level immigration enforcement in a 2016 study analyzing the consequences of section 287(g) on food security of Latinx immigrant children, finding that enforcement of 287(g) was associated with a 10% increase in food insecurity in Mexican non-citizen households with children [30].

Impact on mental health
A 1996 study examining the effects of Proposition 187 on the use of mental health services in San Francisco reported a 26% decrease in Latinx persons aged 18 to 45 seeking outpatient mental health services, net of any weekly outpatient episodes or shared variance between young Latinx individuals and non-Hispanic whites. Moreover, there was a temporary increase in the use of crisis services by young Latinx individuals in the six weeks following Proposition 187's passage [31].
A 2010 analysis examining the effect of detention and deportation on Latinx families and children in the wake of the Antiterrorism and Effective Death Penalty Act, the IIRIA, and the PATRIOT Act found that higher levels of legal vulnerability corresponded to a greater impact of detention or deportation on several critical aspects of their lives, including parent emotional well-being, ability to provide financially, relationships with children, children's emotional well-being and academic performance, and ultimately, poor outcomes for children [32]. A study assessing the health and mental health effects of Section 287(g) of the INA and the Secure Communities Program on mixed-status households identified a causal relationship between local immigration enforcement policies and adverse effects on the health and mental health status of mixed-status Latinx households [33].
Three studies assessed the effects of anti-immigrant policies on the mental health of Latinx children, two of which focused on SB1070. These studies found that SB1070 had a negative emotional impact on children whose parents had been deported or who feared deportation, and ultimately support emotional trauma as a serious unintended consequence of anti-immigrant policies. Analysis of youth behavior in one Arizona middle school revealed that awareness of SB1070 may adversely impact academic adjustment and regulatory behavior in the classroom, which may in turn diminish education achievement; males who were aware of SB1070 reported lower regulatory behavior at a later time point [34,35]. An evaluation of the mental health effects of Attorney General Sessions' 2017 and 2018 directives that dramatically increased deportation rates and intensified a policy of family separation on Latinx adolescents indicated that adolescents with a family member detained or deported had higher odds of suicidal ideation (

Impact on public health and daily life
Two studies found adverse effects of Arizona's SB1070 on self-reported health and increased discrimination, respectively. Spanish-speaking Latinx individuals had a lower predicted probability of excellent and very good health and higher predicted probability of good and fair/poor health (discrete change coefficient of 0.033), with a formal LR test producing significant results. In an assessment of "spillover effects" of SB1070, study participants reported experiencing discrimination in environments including work, school, healthcare, and within their communities. Using a social determinants of health framework, these experiences adversely impact participant economic stability, education, health access, and limit social interactions [37,38].
A 2014 study evaluating changes in Latinx public healthcare service utilization prior to and following Alabama's HB56 found that total monthly visits were at least 20% lower relative to the same month in the previous year, with an overall decrease of 30% relative to the previous year for all service types, including communicable diseases, sexually transmitted infections, and immunizations. For total visits and four of the five specific visit types, the mean monthly percent change differed significantly for Latinx adults vs. non-Latinx adults (p<0.01) [39].
A 2015 study examined the impact of proposed restrictionist immigration ordinances in 24 Pennsylvania counties on local gun ownership. The study included all ordinances proposed by elected officials that directly or indirectly targeted undocumented immigrants. Although none of the anti-immigrant measures was ultimately enacted, their proposal was associated with an increase in handgun sales in these counties, raising concern for a potential increase in gun violence [40].

Discussion
Our findings demonstrate how specific anti-immigrant legislation negatively affects the communities they target. The US is unquestionably experiencing heightened anti-immigrant sentiment. For example, in the Latinx community a 2016 study found that 70% of Latinx individuals reported discrimination in their daily life, a dramatic increase from 30% reported in 2002-2003 [41]. Previously, it has been shown that an anti-immigrant climate is associated with worse health outcomes, poorer self-reported mental health, and increased discrimination in Latinx communities [41][42][43][44][45].
Previously, and predictably environments hostility towards immigrants have been shown to correlate with decreased or delayed healthcare utilization. In California, for example, a main predictor of not seeking medical care is undocumented status [46]. Even with basic health care services that remain covered, such as childhood immunization, the incongruous policy between one government agency aiming to persecute undocumented individuals as another encourages them to seek preventative healthcare services is highly problematic, particularly in the era of the "public charge" which penalizes non-emergent Medicaid usage. Resultant delays in medical diagnoses, linkage to treatment services, and continuation of care for communicable diseases pose a substantial individual and public health risk. Consider, for example, the 1980s Los Angeles measles outbreak, which was partly attributed to low vaccination uptake by the children of undocumented individuals [47,48]. At this time, Latinx two-year old children in California were 50% less likely to be up-to-date with measles immunization than their Non-Hispanic White counterparts, while Latinx children in families with a Mexicanborn parent and child were 15 times more likely to underutilize healthcare and 43 times more likely to be unimmunized to measles than Latinx individuals with a US-born parent and child [49,50].
Even modest deterrents of healthcare uptake can have considerable consequences for public health. For example, a 1994 study by Asch et al. noted that, while only 6% of undocumented patients with active tuberculosis feared that seeking treatment might lead to trouble with immigration authorities, those individuals were almost four times as likely to delay seeking care for more than two months, a period likely to result in disease transmission [51]. Furthermore, it is important to acknowledge that undocumented immigrants are overwhelmingly excluded or discouraged from accessing care, and that new strictures only compound preexisting and longstanding barriers to access including lack of insurance, high mobility, and low health literacy, among others [49]. This conclusion provides considerable reason for concern that contemporary anti-immigrant legislation has more than likely exacerbated the current epidemic we face.
It is worth noting that some policies, while increasing fear and confusion among immigrant communities, had little or no demonstrable effect on healthcare utilization or outcomes [10,11,14,20,24,29]. Several authors credited the inability to select for undocumented individuals, and thus the inadvertent inclusion of permanent residents, refugees, and asylees in their studies as a potential yet important confounder that might underestimate the true harms of the evaluated laws given their continued protection and coverage despite restrictions placed upon undocumented individuals. Additionally, many authors credited the expansion of safety net programs and training of frontline healthcare staff with mitigating potentially devastating health outcomes for these communities. Indeed, particularly in the era of the new "public charge" rule, more information dissemination, engagement, and support at the community level is necessary to maximize health equity for immigrant communities.

Limitations, conclusions, and future directions
Our study has several limitations. First, extensive data heterogeneity precluded meta-analysis. Second, while all immigrant communities were of interest and no restrictions were placed upon our search strategy regarding the race or ethnicity of subjects, the overwhelming majority of studies eligible for screening focused exclusively on Latinx communities, and for that reason our final manuscript retains the same focus as a function of the studies that were ultimately included. Additionally, as previously stated, we acknowledge that Latinx Americans are not all immigrants, and that this study fails to assess the adverse health effects of punitive laws on non-Latinx immigrant communities. Moreover, most studies found it logistically impossible to isolate undocumented individuals for analysis, using proxies such as zip code, level of educational attainment, or uninsured or Spanish-speaking status to estimate their risk. Thus, much of the above data may underestimate the true effects of these laws on undocumented communities.
In conclusion, our study shows that many punitive immigrant policies have decreased immigrant access to and utilization of basic healthcare services, while instilling fear, confusion, and anxiety in these communities. The federal government should preserve and expand access for undocumented individuals without threat of deportation to improve health outcomes for US citizens and noncitizens. Future research on this topic should attempt to analyze the effects of anti-immigrant legislation specifically on undocumented individuals to more accurately assess outcomes and provide stronger evidence to guide policy.