Rabia Ali Khan ,Jo BrownDono Widiatmoko

1College of Health and Social Care,University of Derby,Derby,United Kingdom; 2Joint Modelling Team,Emergency Response Department,Public Health England.

Key words:Influenza vaccination,Asian minority,Self-reported vaccination uptake,Determinants of influenza vaccination,Disparities in vaccination,Minority health

Background

Influenza is a highly contagious acute respiratory infection,estimated to cause 3−5 million cases of severe illness worldwide [1] and is associated with lost productivity and economical costs [2].Compared to other developed nations,influenza associated mortality is evidently much higher in the U.S as a result of the ageing baby boomers [3].The 2017/18 winter season witnessed a record number of 80,000 influenza associated deaths in the U.S.[4,5].

Sub-optimal vaccination levels have been reported in the country despite long standing routine vaccine recommendations [6–8].In 2016−17,only 43.3% of the U.S.adults reported as vaccinated for influenza [9].In contrast with other developed countries,the citizens of the U.S.do not have access to universal health care insurance,however employer provided and privately purchased plans are majorly used to cover healthcare costs [10].

The “Healthy People” program,introduced in 2000 by the Office of Disease Prevention and Health Promotion aims to improve the overall health of the American population through various strategies including increased access to screening programs for atrisk groups and improving immunizations coverage for preventable diseases [11].Since 2010,the Advisory Committee on Immunization Practices,CDC recommends annual influenza vaccination for everyone 6 months and above [12,13].Santibanez et al.(2017)[14] reported a low seasonal influenza vaccination coverage (IVC) amongst adults in the U.S.from 2009−2016.For 2015−16,the national average for influenza immunization uptake amongst adults was reported as 41.7% only [14].These data indicated that the vaccination coverage over the last decade or so fell short of the Healthy People 2020 goals [15].

The Asian ethnic group is increasing at the same rate as the Hispanics and four times faster than the general population in the U.S.[12].As of 2017,more than 22 million individuals of Asian descent resided in the country [16].As of 2020,15.5% of California’s population is comprised of individuals who identify as“Asian alone”.Only 2 studies were identified assessing influenza vaccination disparities amongst Asians in the U.S.[12,17].However,IVC amongst the Asian population of California is under researched and there is an absence of literature on exploring generational status as a predictor for IVC amongst the aforementioned group.Previous research confirms that correlates of health vary by generation [1,18].A recent study indicated generational status as an important determinant for IVC amongst Mexican identified Latino adults [18].Mendiola,Do-Reynoso and Gonzalez(2016) [18] utilized the California Health Interview Survey (CHIS) 2011−12 survey data and reported a lower IVC amongst the third and second-generation Mexicans compared to the first-generation Mexicans.Assessing IVC and exploring generational status as a potential determinant of IVC amongst the Asian population in California will provide valuable insight into the vaccination trends amongst the respective ethnic group.Furthermore,such findings can inform targeted vaccination initiatives aimed to achieve nationally recommended immunization levels as set by the Healthy People 2030 program.

This study aims to report IVC amongst the overall,first and third-generation Asian population in the state of California.We also assessed socio-demographic factors and generational status as independent predictors of IVC amongst the first and third-generation Asians.

Methods

The CHIS is the largest state-wide population-based health survey conducted in the U.S.[19].Since 2001,it has been collecting data on demographics and health care needs of the 58 counties of California [19].The data is de-identified and assigned case number as per UCLA policies [19].It utilizes a multi-stage sampling design and is a random-digit-dial telephone survey available in 6 languages i.e,English,Spanish,Chinese(Mandarin and Cantonese dialect),Tagalog,Vietnamese and Korean [20,21].

The data for this study was drawn from the CHIS 2016 dataset for adults (individuals age 18 + years)conducted between January and December,2016 [19].We considered the CHIS 2016 dataset as the most relevant for our study’s aims because it is the most recent as well as the last year CHIS inquired about IVC[19].The CHIS questionnaires are updated regularly with questions exploring issues of emerging concerns.

Study population

The survey respondents included individuals aged 18 years and above,living in the state of California.Institutionalized individuals and undocumented migrants are excluded from participating in the CHIS[22].Respondents were included based on the following criteria: selected “Yes” for “Received influenza vaccination in the past 12 months” and selected “Asian”for “OMB/Current DOF-Ethnicity” (note: OMB-Office of Management and Budget; DOF-Department of Finance).

Outcome variable

To determine IVC,the survey question “Flu vaccine in past 12 months: Yes/No” was considered as the outcome variable with the selected response “Yes”.

Previously,Mangtani Shah and Roberts (2007) [23],Rolnick et al.,(2013) [24] and Srivastav et al.(2018)[17] have verified self-reported IVC as a reliable and highly sensitive measure of immunization uptake.

Independent variables

The independent variables included gender,age,marital status,location of residence,household size,educational attainment,and the presence of hypertension,diabetes and asthma.These variables have been confirmed as either positive or negative corelates of IVC and have been previously studied amongst various populations in the U.S,Asia and Europe [1,12,25,26].

For generational status,two new variables “First generational status” and “Third generational status”were created by combining selected cases from the“citizenship” and “OMB/DOF-Ethnicity” variables of the CHIS 2016 survey.

Definitions of generational status

Generational status was classified based on the U.S.Census Bureau’s definitions as adopted by Trevelyan et al.(2016) [27].

First-generation: individuals who are not U.S.citizens at birth.This includes naturalized citizens,legal permanent residents and temporary migrants such as foreign-born students.

Second-generation: U.S.born individuals with at least one foreign-born parent.

Third-generation: U.S.born individuals with both U.S.born parents.

Ethical approval

Ethical approval was sought from the Health and Social Care Research Ethics Committee at the university.Given that CHIS is a public database with deidentified data,no ethical approval was required from UCLA for using the data as per the respective institution policies[21].

Statistical methodology

In order to assess IVC,descriptive statistics were utilised,frequencies and percentages were obtained,and comparisons were drawn amongst the overall,first and third-generation Asian respondents.

To explore generational status as a determinant of IVC amongst the Asian ethnic group,weighted binary logistic regression analysis was utilized.Comparisons were made between the first and third-generation Asian groups using adjusted odds ratio (AOR).Data was controlled for “insurance status” and “visit to the doctor”.

Data analysis were conducted using IBM SPSS 25.0 for Mac.A 2 tailedPvalue ＜ 0.05 was determined as statistically significant.

Note: second-generation Asians were excluded from the regression analysis owing to the limitations of the IBM SPSS 25.0 software.We were unable to select the second-generation Asians by filtering the first and thirdgeneration cases from the overall Asian sample.

Results

Of the 21,055 total respondents,13.1% (n = 2761)identified as Asians.The Asian respondents were classified into first,second and third-generation Asian’s as illustrated in Table 1.The highest percentage of the total Asian respondents comprised of the firstgeneration individuals (n = 1,857,67.26%)

Demographic distribution of the Asian sub-groups

Both the first and third-generation Asians comprised of a higher proportion of females than males (1stgeneration: M = 46.4%; F = 53.4% vs 3rdgeneration: M= 46.7%; F = 53.3%) (Table 1,2).Adjusted logistic regression indicates females from the first-generation were more likely (AOR = 1.108,confidence interval(CI): 1.103,1.114) while the third-generation females were less likely (AOR = 0.777,CI: 0.752,0.802) to report IVC (Table 3).When comparing the various age groups,a larger proportion of the first-generation Asians comprised of the 18–39 year old (1stgeneration= 15.8% vs 3rdgeneration = 10.0%) while the thirdgeneration majorly constituted of the 65 years and above age group (1stgeneration = 43.1% vs 3rdgeneration = 47.4%).The 65 + years age group was more likely to report IVC amongst the first and third generations (1stgeneration = AOR = 5.656,CI: 5.603,5.710) (3rd generation = 2.821,CI: 2.691,2.958) to report IVC.

Table 1 Asian respondents identified in CHIS 2016 stratified by generational status (Source: California Health Interview Survey 2016)

Table 2 Population characteristics of the total,first and third generation Asian respondents

A higher proportion of the respondents from the firstgeneration were married (62.6%) compared to the thirdgeneration respondents (48.1%).Adjusted logistic regression indicates that amongst the first-generation,married respondents were 1.283 (CI: 1.276,1.291)times likely to report IVC compared to 0.661 (CI: 0.635,0.688) times amongst the third-generation Asians.Both the first and third-generation majorly lived in urban settings (1stgeneration = 97.3 vs 3rdgeneration = 94.8).Rural inhabitants were 2.824 (CI: 2.768,2.882) times more likely to report IVC.

Higher proportion of the first-generation Asians lived in households of size 3 and above (44.0%) while the highest proportion of the third generations lived in household of size 1 (37.7%) followed by households of size 2 (36.7%).AOR indicated that amongst the firstgeneration,respondents from household of size 3 were the least likely to be immunized (AOR: 0.546,CI:0.540,0.553) while this appeared as an insignificant predictor amongst the third-generation.Though in both the groups,highest proportion of individuals had obtained a bachelor’s degree (1stgeneration = 33.1 vs 3rdgeneration = 43.6),the third-generation consisted ofalmost 10% more respondents with a bachelor degree compared to the first-generation.Comparatively,the first-generation had a higher percentage of individuals with no formal education/grade 1–8 (1stgeneration =6.8 % vs 3rdgeneration = 0.3%) and high school diploma (1stgeneration = 19.0% Vs.3rdgeneration =9.0%).AOR for the education variable indicate that first-generation Asians with a university qualification are 1.573 (CI:1.565,1.582) times more likely,while the third-generation with a university qualification are 2.394 (CI 2.302,2.488) times more likely to report IVC.

Table 3 AOR for the first and third generation Asians (controlled for insurance status and visit to the doctor)

Majority of the first and third-generation Asians had insurance coverage (1stgeneration = 95.7% vs.3rdgeneration = 98.6%).A higher proportion (88.6%) of the third-generation respondent confirmed visiting the doctor in the past twelve months compared to only 82.2 % of the first-generation respondents.Additionally,the third-generation also constituted a higher percentage of hypertensive (48.8% vs 40.7) and asthmatic (17.6% vs 8.2%) respondents.The difference in confirmed diabetes amongst the 2 population was not as pronounced as hypertension and asthma with 18.3%of the first-generation and 20.4% of the third-generation reporting a diagnosis of diabetes.However,AOR indicates the third-generation Asian respondents with diabetes as 73.930 (CI: 66.138,82.642) times likely to report IVC.The first-generation diabetic respondents were only 2.120 (CI: 2103,2.138) times likely to report vaccination but not as distinct as the third-generation.Having a diagnosis of hypertension was associated with a reduced likelihood of reporting IVC amongst both the generations (1stgeneration AOR: 0.656 vs 3rdgeneration AOR: 0.357),with significantly lower amongst the third-generation.Third-generation Asians with asthma were more likely to report IVC compared to first generation (AOR 3rd generation: 2.863 vs AOR 1st generation: 1.377).

When assessed via weighted binary logistic regression and controlled for “insurance status” and“visit to the doctor”,all predictors appeared statistically significant (P＜0.01) amongst the first-generation,while location and household size were insignificantly corelated amongst the third-generation Asians.

Discussion

Our findings indicate that females are more likely to report IVC amongst the Asian ethnic group.This also accords with previous observations by Vaidya,Partha and Karmakar (2012) [28] and Almario et al.(2016) [6].Green and Pope (1999) [29] had commented that compared to males,females were more sensitive to illness and focused more on preventive health.Females have also been reported to have a higher number of routine doctor visits [30] and studies have consistently reported visits to the doctor as positively correlated with an increased IVC [4,31].However,we confirm females as less likely to report IVC amongst the third generation Asians.This is incoherent with previous literature which reported females as more likely to report IVC amongst third-generation Hispanics [18] and owing to limited literature and qualitative studies we are unable to ascribe possible reasons to our findings.

Majority (61.1%) of the Asian’s in California are foreign born.Previous studies have attributed the“Healthy Migrant Effect” which asserts that foreign born individuals are healthier and less risk aversive compared to American born individuals,which may explain their high utilization of preventive care [12,32].Chung et al.(2018) [33] confirmed that compared to Whites,Asians had the highest odds of preventive healthcare visit (AOR: 0.853 (African-American),0.886 (Hispanic),and 1.236 (Asian)).

However,when comparing AOR vales amongst the first and third-generations we noticed interesting and surprising outcomes.Our data illustrates that the thirdgeneration Asians are more likely (70.9%) than firstgeneration (61.1%) to report IVC.This is similar to the findings of Mendiola,Do-Reynoso and Gonzalez’s(2016) [18] study exploring IVC in U.S.born and foreign-born Hispanics.Though they reported a similar trend,their IVC proportions were notably lower (1stgeneration = 26% vs 3rdgeneration = 33%) which could be reflective of the overall low IVC amongst Hispanics.The higher IVC evident in the third-generation Asians may be a result of the sample distribution since the group pre-dominantly consisted of individuals 65 +years,while the first-generation consisted of a higher proportion of 18–39 year old’s.When comparing the AOR values for the predictors and IVC amongst the first and third generation,our results for gender and marital status considerably varied from Mendiola,Do-Reynoso and Gonzalez’s (2016) [18] study.As indicated,the trends are similar (both decreasing from the first to third generation),but the reduction of the AOR values is significantly different.This indicates that Hispanics and Asians have unique behaviors and hence require health promotion strategies unique to their respective ethnic groups.

Majority of the Asian population reported educational attainment of a bachelor’s degree or higher.This eventually translates into better employment opportunities and increased monthly income.These factors together could possibly attribute to the increased IVC evident from this group.Jain et al.(2017) [34]confirmed that in countries where payment of vaccination is necessary,individuals with higher education had a 67% increased odds of vaccination uptake.

Limitations

Though offered in 6 languages,the CHIS is not available in any of the South-Asian languages,a subgroup confirmed to be rapidly growing in the U.S.Hence South-Asians preferring to speak in languages other than English,may be underrepresented.Also,the sample only included respondents from California,a state majorly inhabited by ethnic minorities and hence our findings may not be generalizable to states with dissimilar demographics.

Additionally,since Asians comprise of diverse ethnic groups,aggregating them together may mask the differences prevalent amongst each group respectively.Hence a generalized classification is conceptually problematic,and interventions developed based on these findings may not yield maximum results.In order to improve the outcomes of public health vaccination programs,nuanced targeted health policies are required as opposed to broad-based interventions.

Conclusion

The Asian minority reported a higher IVC compared to the overall population of California and the U.S.A.One of the more significant findings to emerge from this study is the high rate of vaccination reported by the third-generation Asians.This is a new finding and supplements current literature which has till date only explored generational status amongst Mexican identified Latino adults only.The predictors for the first and third-generation Asians were quite distinct,with females more likely to report IVC amongst the first generation while males were more likely to report IVC amongst the third-generation.Likewise,being married was associated with an increased vaccination uptake amongst the first-generation while being unmarried was positively associated with increased uptake amongst the third-generation.Hence,the striking diversity amongst the two generations raise important concerns over the current immunization programs which attend to Asians as a homogenous group rather than stratified by generation.