Use of the Improvement Index to Evaluate Equitable COVID-19 Vaccine Allocation in the San Francisco Bay Area

Riana B Jumamil; George Rutherford

doi:10.1016/j.focus.2023.100075

. 2023 Feb 4;2(2):100075. doi: 10.1016/j.focus.2023.100075

Use of the Improvement Index to Evaluate Equitable COVID-19 Vaccine Allocation in the San Francisco Bay Area

Riana B Jumamil ^1,^⁎, George Rutherford ²

PMCID: PMC9897871 PMID: 36776788

HIGHLIGHTS

•
A social determinants of health index can help target public health interventions.
•
The improvement index has advantages over conventional health equity metrics.
•
California's strategy using a social determinant of health index led to equitable COVID-19 vaccine allocation.

Keywords: COVID-19, vaccination, public health intervention, health disparity, San Francisco Bay Area, improvement index

Abstract

Introduction

A social determinants of health index score or Vaccine Equity Metric was used to prioritize resources and address geographic disparities in California's vaccination coverage. We calculated the improvement index or percentage of the vaccination disparity gap closed to evaluate the impacts of this vaccination strategy in the San Francisco Bay Area during the SARS-CoV-2 Delta variant surge.

Methods

We conducted a cross-sectional study on San Francisco Bay Area ZIP codes during the Delta surge (July 6–October 5, 2021). Data came from the California Immunization Registry and the 2019 5-year American Community Survey. We used Spearman correlations to examine the relationships between Vaccine Equity Metric category and vaccine coverage and Kruskal–Wallis tests to compare vaccination improvement index across Vaccine Equity Metric categories.

Results

We studied 248 ZIP codes in the San Francisco Bay Area. Those with the lowest resources (Vaccine Equity Metric Level 1) had the highest absolute increase in vaccination coverage (14.3 vs 5.4 percentage points in Vaccine Equity Metric Level 4), although a contribution was higher starting vaccination rates in Level 4 ZIP codes with the greatest resources. The ratio of vaccination coverage between the lowest- and highest-resourced ZIP codes increased from 0.79 to 0.9, suggesting reduced disparity. However, it is difficult to interpret given wide differences in n (Level 1 n=8 vs Level 4 n=151). In contrast, the vaccination improvement index accounts for each Vaccine Equity Metric category's baseline vaccination; all were statistically similar (grand mean=41.5%, p=0.367), implying comparable improvement across all ZIP codes.

Conclusions

Using a Vaccine Equity Metric to identify and prioritize resources to vulnerable communities contributed to equitable vaccine allocation in the San Francisco Bay Area. Our study shows an example of the improvement index's advantages over conventional health equity metrics, such as absolute differences and relative effect measures, which can overestimate an intervention's impact.

Graphical Abstract

INTRODUCTION

In the summer of 2021, the Delta variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) became the dominant strain in the U.S.¹ During this time, the Centers for Disease Control and Prevention recognized the San Francisco Bay Area as a hot spot on July 6, 2021, with rapidly increasing coronavirus disease 2019 (COVID-19) cases as well as hospitalizations and deaths.² Although several Bay Area counties had higher vaccination coverage than the then national average of 64% in fall 2021, there were predictably large discrepancies between highly affluent and less affluent areas. Higher burdens of disease and mortality rates prevailed among communities with lower vaccination rates, which are greatly impacted by systemic inequities such as structural racism.3, 4, 5 Historic redlining and disinvestment have been associated with lower income in minority communities as well as poorer physical and mental health outcomes.⁶^,⁷ Lower vaccination in these areas may be a contribution to decreased access to health care, health insurance, transportation, education, and limited English proficiency.8, 9, 10, 11 Furthermore, initial vaccine hesitancy was prevalent among Black and Hispanic communities owing to the previous experience of racial discrimination, medical distrust, and fear of vaccine side effects.⁸^,⁹ Therefore, there have been geographic inequities in vaccine access, with ZIP codes with predominantly higher income and White populations having higher vaccination rates and lower COVID-19 mortality.⁴^,⁵

Building an equitable vaccination strategy should include practices and policies that expand the supply and access to vaccines in high-risk communities.¹²^,¹³ A helpful measure to identify these communities may be a calculated social determinants of health (SDOH) index score, which can help public health agencies determine where to prioritize resources.¹⁴ An SDOH index score considers several social factors (such as income, housing, and health care) that contribute to health outcomes and can provide insight into an individual and community's well-being. Moreover, health inequities have been traditionally reported in absolute or relative effect measures, which carry limitations when assessing public health interventions.¹⁵^,¹⁶ Relative effects measures, such as risk ratios and ORs, are based on probabilities and may have a large magnitude even though the actual public health impact is minimal. Absolute effect measures, such as attributable risk or risk differences, can reveal initial inequities and where improvement is needed. However, it is difficult to assess change between populations that have different starting values. We chose to examine the improvement index (II) or percentage of the vaccination gap closed between different ZIP codes categorized by SDOH index score during the Delta wave.¹⁶ The II, as described by Katuramu et al.,¹⁶ is a useful metric because it can compare improvement across categories despite different baseline vaccination numbers (because ZIP codes with already highly vaccinated populations do not have much room for improvement). Therefore, a similar II across SDOH categories may be an outcome that better implies equitable vaccine allocation. Ultimately, more resources must be prioritized toward areas where greater inequities exist to achieve such an equitable outcome.

In our study, we noted that the California Department of Public Health (CDPH) used a ZIP code–assigned SDOH index score to conduct an equitable vaccine allocation strategy and public health outreach to high-risk communities.¹⁷ By evaluating the II, we explored whether the CDPH's strategy was useful in increasing COVID-19 vaccination coverage among previously unvaccinated Bay Area residents across all SDOH categories during the Delta surge. The findings of our study can inform future strategies as new variants of COVID-19 arise.

METHODS

Study Population

We conducted a cross-sectional study of San Francisco Bay Area ZIP codes, defined as ZIP code tabulation areas, during the Delta variant surge from July 6 to October 5, 2021. We included ZIP codes in Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, and Sonoma counties. We obtained publicly available data on COVID-19 ZIP code vaccination rates and reported cases of COVID-19 from the CDPH and California Immunization Registry (CAIR) during the Delta surge period.¹⁸ We also obtained total population size estimates and demographic characteristics of ZIP codes from the 2014–2019 American Community Survey (ACS), which represents U.S. Census Bureau data estimated over 5 years.¹⁹ Data included household residents who were aged ≥12 years (for whom COVID-19 vaccines were approved at the time).

Measures

The Vaccine Equity Metric (VEM) combines CDPH-derived scores with the California Healthy Places Index (HPI), a health index score developed by the Public Health Alliance of Southern California that is validated against life expectancy at birth as measured by publicly available census data.¹⁴^,¹⁷ The HPI is a census tract–level index score that combines factors (income, education, transportation, housing, health care) that contribute to a healthy community and are associated with decreased COVID-19 mortality. It is comprised weighted scores in 8 domains: economic resources, social resources, education, transportation, neighborhood, housing, clean environment, and healthcare access. ZIP codes without an HPI score included those with a population <1,500 or where >50% of the population resides in group quarters owing to concerns with statistical reliability and validity in smaller populations. These ZIP codes were assigned a CDPH-derived score that attributed similar HPI criteria to create an SDOH index rating. VEM scores for all California ZIP codes were categorized by quartile into 4 categories. ZIP codes ranged from least healthy community conditions in Level 1 (lowest quartile) to most healthy community conditions in Level 4 (highest quartile).

Since March 2021, the CDPH has used VEM category to target its equitable vaccine allocation and access strategy.¹⁷ This was imperative because socioeconomically deprived areas were shown to have lower vaccination coverage and suffer from higher COVID-19 case rates and mortality.⁴^,⁵^,¹⁰ In May 2021, 40% of the state's supply of doses were dedicated to the lowest VEM quartile to improve vaccination equity. The CDPH had evidence-based efforts to combat the challenges of decreased healthcare access, income, and transportation in these areas. It moved away from mass vaccination sites and toward smaller neighborhood vaccination clinics that were coordinated with and run by trusted sources, including community-based organizations, places of worship, businesses, and schools.20, 21, 22 They also provided free transportation to vaccination sites as well as a home vaccination program for homebound residents. Over $85 million were supplied to local organizations to perform community outreach, including neighborhood canvassing programs to schedule vaccine appointments.

VEM category is the main predictor variable in this study and our unit to an analysis by ZIP code. We excluded ZIP codes from our study if they did not have a calculated VEM category, if they were not from the Bay Area, and if 100% of the area was vaccinated by July 6.

Our primary outcome was the II or mean percentage increase in new vaccine doses between July 6 and October 5, 2021 among residents who were unvaccinated on July 6, 2021. Data were retrieved from the CAIR, which recorded daily vaccinations. We obtained the percentage of the first COVID-19 vaccine doses among Bay Area ZIP codes on July 6, 2021. The number of unvaccinated residents at the start of the Delta wave was estimated by subtracting the fraction of residents with at least 1 vaccine dose (A) from 1 and multiplying the difference by the ACS-estimated total ZIP code population aged ≥12 years (B), as in the following equation: (1—A) $\times$ B. To calculate the percentage increase in vaccination, we took the difference in unvaccinated residents from July 6 (C) and October 5 (D) and divided it by the number of vaccinated residents on July 6, as in the equation: $\frac{C - - D}{C} \times 100$ .

We examined the relationships between VEM category and ZIP code sociodemographic characteristics not inherent within the VEM metric. These characteristics were calculated as the population percentage of White, Black or African American, Asian, and Indigenous (including Native Hawaiian, Pacific Islander, and Alaskan Indian) race; of Hispanic or Latinx ethnicity; aged ≥65 years; of male sex; and of non-U.S. citizenship. In addition, we explored the associations between the ZIP code characteristics mentioned earlier and vaccination coverage before and after the Delta variant surge.

Statistical Analysis

We used an independent-samples Kruskal–Wallis test to compare nonparametric continuous variables among VEM categories, such as ZIP code vaccination II. We calculated Spearman correlation to determine whether VEM category was significantly associated with vaccination coverage before and after the Delta variant surge. We also used Spearman correlation to determine a relationship between VEM category and the percentage population of race and ethnicity, age, sex, and non-U.S. citizen status.

We then analyzed ZIP codes in the highest VEM category (Level 4) or the healthiest communities separately from ZIP codes in the lowest VEM categories (Levels 1–3) or the less healthy communities to assess differences. We calculated a multivariate linear regression assessing the relationship between ZIP code sociodemographic characteristics (race, ethnicity, age, sex, and non-U.S. citizenship) and vaccination II during the Delta surge. Statistical significance was defined as 2-sided alpha at 0.05.

RESULTS

We investigated 248 ZIP codes in the San Francisco Bay Area. The majority had high healthy community index scores, with 8 in the first, 33 in the second, 56 in the third, and 151 in the fourth VEM category levels. The median (IQR) vaccination percentage with at least 1 COVID-19 vaccine was 81.8% (IQR=15.0 percentage points) on July 6, 2021 and 89.2% (IQR=12.3 percentage points) on October 5, 2021 at the end of the Delta variant wave.

Table 1 illustrates ZIP code characteristics as a median percentage across VEM category. ZIP codes in the lowest VEM category (Level 1), the most socioeconomically deprived areas, had statistically significantly lower median vaccination percentages than in the highest health category (Level 4) at the beginning (68.3% vs 86.3%) and end (82.6% vs 91.7%) of the Delta wave by Spearman correlation (p<0.001). This corresponded to an indirect relationship between VEM level and COVID-19 case rates at the end of the Delta wave, with VEM Level 1 ZIP codes having the highest case rates at 228 per 10,000 residents compared with 140 per 10,000 residents among VEM Level 4 ZIP codes. High VEM category levels were associated with ZIP codes having a large White race population (Spearman 0.312, p<0.001) and low Black or African American race population (Spearman –0.373, p<0.001). There was a borderline significant relationship between increasing VEM levels and population percentage of the Asian race (Spearman 0.125, p=0.05). A lower population percentage of Latinx/Hispanic ethnicity was associated with a higher VEM category (Spearman –0.187, p=0.003). ZIP codes in high VEM categories also had a larger population of elderly people aged ≥65 years (Spearman 0.220, p<0.001) and a lower population of people who are not U.S. citizens (Spearman –0.274, p<.001).

Table 1.

ZIP Code Characteristics by VEM Category

	VEM level^b
ZIP code characteristics^a	First (n=8)	Second (n=33)	Third (n=56)	Fourth (n=151)
Percentage of adults who had received at least 1 vaccination dose on July 6, 2021^⁎⁎⁎	68.3	73.3	76.0	86.3
Percentage of adults who had received at least 1 vaccination dose on October 5, 2021^⁎⁎⁎	82.6	84.8	84.3	91.7
COVID-19 cases per 10,000 population on October 5, 2021^⁎⁎⁎	228.0	215.5	185.5	140.4
Race, %
Asian	10.5	15.0	16.3	17.8
Black or African American^⁎⁎⁎	20.8	5.2	3.5	2.0
Indigenous^c	1.1	0.9	0.8	0.8
White^⁎⁎⁎	37.7	39.2	51.3	63.9
Ethnicity, %
Hispanic or Latinx^⁎⁎	18.1	18.9	21.7	15.0
Age ≥65 years, %^⁎⁎⁎	8.7	12.0	15.4	16.4
Not a U.S. citizen, %^⁎⁎⁎	21.5	17.9	11.8	9.2

Open in a new tab

Note: Asterisks indicate statistically significant differences by category (*p<0.05, ^⁎⁎p<0.01, and ^⁎⁎⁎p<0.001).

ZIP codes are defined as ZCTAs.

Unless otherwise noted, each cell in this table reports a median percentage for neighborhoods in the given VEM category. A VEM estimates factors that impact health such as income, education, health insurance status, and access to transportation. They combine the Public Health Alliance of Southern California's HPI measure with scores derived from the California Department of Public Health. ZIP codes range from less healthy community characteristics in the VEM Level 1 (first quartile) to more healthy community characteristics in VEM Level 4 (fourth quartile).

Indigenous is defined as American Indian, Alaska Native, Native Hawaiian, and another Pacific Islander.

HPI, Healthy Places Index; VEM, Vaccine Equity Metric; ZCTA, ZIP code tabulation area.

Virtually all Bay Area ZIP codes experienced an increase in vaccination coverage during the Delta wave (Figure 1). ZIP codes in the lowest VEM categories (Levels 1 and 2) had the highest absolute increase in vaccination coverage (14.3 and 11.5 percentage points, respectively) than those in Levels 3 and 4 (8.3 and 5.4 percentage points, respectively). The ratio of vaccination coverage between the lowest (Level 1) and highest (Level 4) resourced ZIP codes increased from 0.79 on July 6 to 0.9 on October 5, suggesting a decrease in disparities at the end of the Delta surge (where a ratio of 1 would imply no difference). The vaccination II or mean percentage increase in vaccination coverage among all the 4 VEM categories was statistically similar by Kruskal–Wallis test (grand mean=41.5%, p=0.367) (Figure 2). The average II among VEM Levels 1–4 was comparable (46.9%, 43.4%, 38.1%, and 42.0%, respectively), implying at least an equal improvement in vaccination coverage, despite marked absolute and relative effect findings.

SF Bay Area ZIP code vaccination coverage by VEM category.

SF, San Francisco; VEM, Vaccine Equity Metric.

Improvement index or mean percentage increase in vaccination coverage among SF Bay Area ZIP codes by VEM category from July 6 to October 5, 2021.

SF, San Francisco; VEM, Vaccine Equity Metric.

In a multivariate linear regression model among all ZIP codes, those with a higher percentage Asian population had higher vaccination coverage both before and after the Delta surge (0.004, p<0.001 and 0.002, p<0.05), as did those with a higher percentage elderly population (0.003, p<0.01 and 0.003, p<0.05, respectively). With regard to percentage increase in vaccinations, a higher White population was associated with decreased vaccination II among ZIP codes in VEM Levels 1 through 3 (–0.005, p=0.03). Among ZIP codes in the healthiest VEM category (Level 4), racial, ethnic, elderly, sex, and non-U.S. citizen population were not associated with vaccine improvement.

DISCUSSION

There are many ways to define health equity—it may mean equal healthcare access, equal utilization of health care, resource allocation according to need, or equal health outcomes, among others.²³ As such, there are different metrics to target health equity, each with particular insights into a situation. Traditional measures of equity and an intervention's impact have included absolute risk differences, RR effects, and number needed to treat.¹⁵ Risk differences and number needed to treat are helpful to inform which groups have the highest need and whether an intervention has increased utilization of health care. RR can be used to compare health outcomes across groups and be used to target equal outcomes. Another metric may be targeting a health outcome above a certain threshold for all groups, which is a specific target often used in quality improvement, although the threshold point is often arbitrarily set. We discuss the II, which measures the percentage of the disparity gap closed at the end of an intervention and can better reflect expanded access even though disparities may still exist. Public health interventions must recognize the systemic barriers that have led to these inequities to appropriately address them.⁶

In our study, the CDPH prioritized resources toward areas of highest need and lower baseline vaccine coverage before the Delta surge. This resulted in larger absolute increases in vaccination coverage in ZIP codes with the lowest VEM but similar increases in the II across all VEM categories. Had the target been similar increases in absolute vaccinations for all ZIP codes, disparities among VEM categories would still persist. Furthermore, looking at absolute risk differences alone overestimates the effect of the CDPH vaccination strategy because wealthier ZIP codes had higher initial vaccination rates at the start of the Delta surge. Rather, our primary outcome was the II or mean percentage increase in new vaccine doses among previously unvaccinated residents and is another way to view equitable vaccine allocation. The II among all VEM categories was statistically similar, suggesting comparable vaccination improvement despite persisting (although decreasing) disparities at the end of the Delta surge. This provides an example of how selecting the appropriate metric can help to guide more equitable action and designate resources to where they are needed most.

Our sociodemographic findings correlate with existing literature. ZIP codes in the highest VEM category had significantly lower minority populations, higher vaccination coverage, and lower overall COVID-19 case rates. These areas had no associations between vaccination coverage and race, ethnicity, sex, age >65 years, or U.S. citizenship, suggesting that favorable socioeconomic conditions were conducive to vaccine access and acceptance. For all ZIP codes, a higher percentage population of Asian and elderly persons aged ≥65 years was associated with higher vaccination coverage before and after the Delta surge, suggesting early adoption by self-perceived high-risk groups. Sources on COVID-19 U.S. vaccine hesitancy suggest that it is not uniformly distributed by racial group but rather reflects the conservative-leaning ideology and lower perceived threat of infection and structural barriers such as lower income and lack of child care.¹¹ These findings highlight the importance of considering SDOH in planning public health interventions.

Limitations

Our study had several limitations. One consideration is that ZIP code data are only available as a compilation of 5-year ZIP code tabulation area–level estimates. The most recent data are from 2014 to 2019 and do not overlap with the current COVID-19 pandemic, which began in February 2020 in the Bay Area. Although the data are less current, the margins of error are smaller than 1-year estimates owing to larger sample size. Because of increased frequency of sampling, the estimates can be imprecise. Another consideration is that not all U.S. residents are recorded in the ACS. For instance, people experiencing homelessness are not included owing to having no address. We did not have access to estimates of the number of homeless individuals by San Francisco Bay Area ZIP code. Disaggregated data on Asian subgroups were also not recorded in the ACS or CDPH data, so we were unable to elucidate further specific findings. In addition, the CAIR database does not include vaccines administered by the Veterans Health Administration, the Indian Health Service, the Department of Defense, or the Federal Bureau of Prisons. Therefore, residents who received their first vaccine dose through these organizations were excluded.

CONCLUSIONS

The CDPH's use of an SDOH index score is an example of efforts to address the structural barriers formed by poverty and institutionalized racism to create equitable access to vaccines.⁶ Future variants of COVID-19 will continue to lead to high case rates and hospitalizations; therefore, prioritization of socioeconomically disadvantaged areas is needed to allocate public health resources. Consideration of appropriate metrics, which can include the II, is paramount to evaluating public health interventions and achieving health equity.

ACKNOWLEDGMENTS

The authors would like to acknowledge Dr. Gabriela Reed for her contributions to the literature review of this study; Dr. Scott Bauer, Dr. Jeffrey Kohlwes, Dr. Melissa Medvedev, all faculty of the Designing Clinical Research course at the University of California San Francisco (UCSF), and Dr. Isabel Elaine Allen with the UCSF Clinical and Translational Science Institute for their guidance; and the California Department of Public Health.

The contents of this paper are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

This project was supported by the National Center for Advancing Translational Sciences, NIH through UCSF Clinical and Translational Science Institute Grant Number UL1 TR001872. This project was supported by the UCSF Department of Medicine.

Declaration of interest: none.

CREDIT AUTHOR STATEMENT

Riana B. Jumamil: Conceptualization, Formal analysis, Methodology, Validation, Visualization, Writing – original draft. George Rutherford: Data curation, Resources, Supervision, Writing – review & editing.

REFERENCES

1.Bolze A, Luo S, White S, et al. SARS-CoV-2 variant Delta rapidly displaced variant Alpha in the United States and led to higher viral loads. Cell Rep Med. 2022;3(3) doi: 10.1016/j.xcrm.2022.100564. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Larsen K. Bay area lands on CDC’s emerging COVID hot spot list as cases surge among unvaccinated. Abc7 News. July 12, 2021 https://abc7news.com/covid-hot-spot-cdc-spots-coronavirus-bay-are-19-in-california/10876198/ Accessed September 16, 2021. [Google Scholar]
3.Cuadros DF, Miller FD, Awad S, Coule P, MacKinnon NJ. Analysis of vaccination rates and new COVID-19 infections by U.S. county, July–August 2021. JAMA Netw Open. 2022;5(2) doi: 10.1001/jamanetworkopen.2021.47915. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Zeng S, Pelzer KM, Gibbons RD, Peek ME, Parker WF. Association of ZIP code vaccination rate with COVID-19 mortality in Chicago, Illinois. JAMA Netw Open. 2022;5(5) doi: 10.1001/jamanetworkopen.2022.14753. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Sacarny A, Daw JR. Inequities in COVID-19 vaccination rates in the 9 largest U.S. cities. JAMA Health Forum. 2021;2(9) doi: 10.1001/jamahealthforum.2021.2415. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Jones C, The impact of racism on health. Ethn Dis. 2002;12(1):10–13. http://www.jstor.org/stable/45410796. Accessed November 21, 2022. [PubMed]
7.Lynch EE, Malcoe LH, Laurent SE, Richardson J, Mitchell BC, Meier HCS. The legacy of structural racism: associations between historic redlining, current mortgage lending, and health. SSM Popul Health. 2021;14 doi: 10.1016/j.ssmph.2021.100793. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Tram KH, Saeed S, Bradley C, et al. Deliberation, dissent, and distrust: understanding distinct drivers of coronavirus disease 2019 vaccine hesitancy in the United States. Clin Infect Dis. 2022;74(8):1429–1441. doi: 10.1093/cid/ciab633. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Kim D. Associations of race/ethnicity and socioeconomic factors with vaccination among US adults during the COVID-19 pandemic, January to March 2021. Prev Med Rep. 2023;31 doi: 10.1016/j.pmedr.2022.102021. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Lindemer E, Choudhary M, Donadio G, Pawlowski C, Soundararajan V. Counties with lower insurance coverage are associated with both slower vaccine roll out and higher COVID-19 incidence across the United States. Vaccines (Basel) 2021;9(9):973. doi: 10.1101/2021.03.24.21254270. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.El-Mohandes A, White TM, Wyka K, et al. COVID-19 vaccine acceptance among adults in four major US metropolitan areas and nationwide. Sci Rep. 2021;11(1):21844. doi: 10.1038/s41598-021-00794-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Skolnik A, Bhatti A, Larson A, Mitrovich R. Silent consequences of COVID-19: why it's critical to recover routine vaccination rates through equitable vaccine policies and practices. Ann Fam Med. 2021;19(6):527–531. doi: 10.1370/afm.2730. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Kipnis P, Soltesz L, Escobar GJ, Myers L, Liu VX. Evaluation of vaccination strategies to compare efficient and equitable vaccine allocation by race and ethnicity across time. JAMA Health Forum. 2021;2(8) doi: 10.1001/jamahealthforum.2021.2095. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Maizlish N, Delaney T, Dowling H, et al. California healthy places index: frames matter. Public Health Rep. 2019;134(4):354–362. doi: 10.1177/0033354919849882. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.King NB, Harper S, Young ME. Use of relative and absolute effect measures in reporting health inequalities: structured review. BMJ. 2012;345(7878):e5774. doi: 10.1136/bmj.e5774. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Katuramu R, Wallenta J, Semitala FC, et al. Closing the gap: a novel metric of change in performance. East Afr J Appl Health Monitor Eval. 2019;2019(3) http://eajahme.com/wp-content/uploads/2019/02/Katuramu_FINAL.pdf [PMC free article] [PubMed] [Google Scholar]
17.Blueprint for a safer economy. California Department of Public Health. https://www.cdph.ca.gov/Programs/CID/DCDC/Pages/COVID-19/COVID19CountyMonitoringOverview.aspx. Updated June 15, 2021. Accessed November 17, 2022.
18.Vaccination data. CA.gov. https://covid19.ca.gov/vaccination-progress-data/#equitably-across-groups. Updated February 16, 2023. Accessed November 22, 2021.
19.2015-2019 ACS 5-year Estimates. U.S. Census Bureau. https://www.census.gov/programs-surveys/acs/technical-documentation/table-and-geography-changes/2019/5-year.html. Updated December 8, 2021. Accessed October 22, 2021.
20.As California surpasses 30 million vaccines, Governor Newsom doubles down on efforts to vaccinate hard-to-reach communities. Sacramento, CA: CA.gov. https://www.gov.ca.gov/2021/05/04/as-california-surpasses-30-million-vaccines-governor-newsom-doubles-down-on-efforts-to-vaccinate-hard-to-reach-commun. Published May 4, 2021. Accessed January 21, 2022.
21.Barry V, Dasgupta S, Weller DL, et al. Patterns in COVID-19 vaccination coverage, by social vulnerability and urbanicity — United States, December 14, 2020–May 1, 2021. MMWR Morb Mortal Wkly Rep. 2021;70(22):818–824. doi: 10.15585/mmwr.mm7022e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Batteux E, Mills F, Jones LF, Symons C, Weston D. The effectiveness of interventions for increasing COVID-19 vaccine uptake: a systematic review. Vaccines (Basel) 2022;10(3):386. doi: 10.3390/vaccines10030386. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Braveman P. Health disparities and health equity: concepts and measurement. Annu Rev Public Health. 2006;27:167–194. doi: 10.1146/annurev.publhealth.27.021405.102103. [DOI] [PubMed] [Google Scholar]

[bib0001] 1.Bolze A, Luo S, White S, et al. SARS-CoV-2 variant Delta rapidly displaced variant Alpha in the United States and led to higher viral loads. Cell Rep Med. 2022;3(3) doi: 10.1016/j.xcrm.2022.100564. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0002] 2.Larsen K. Bay area lands on CDC’s emerging COVID hot spot list as cases surge among unvaccinated. Abc7 News. July 12, 2021 https://abc7news.com/covid-hot-spot-cdc-spots-coronavirus-bay-are-19-in-california/10876198/ Accessed September 16, 2021. [Google Scholar]

[bib0003] 3.Cuadros DF, Miller FD, Awad S, Coule P, MacKinnon NJ. Analysis of vaccination rates and new COVID-19 infections by U.S. county, July–August 2021. JAMA Netw Open. 2022;5(2) doi: 10.1001/jamanetworkopen.2021.47915. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0004] 4.Zeng S, Pelzer KM, Gibbons RD, Peek ME, Parker WF. Association of ZIP code vaccination rate with COVID-19 mortality in Chicago, Illinois. JAMA Netw Open. 2022;5(5) doi: 10.1001/jamanetworkopen.2022.14753. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0005] 5.Sacarny A, Daw JR. Inequities in COVID-19 vaccination rates in the 9 largest U.S. cities. JAMA Health Forum. 2021;2(9) doi: 10.1001/jamahealthforum.2021.2415. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0006] 6.Jones C, The impact of racism on health. Ethn Dis. 2002;12(1):10–13. http://www.jstor.org/stable/45410796. Accessed November 21, 2022. [PubMed]

[bib0007] 7.Lynch EE, Malcoe LH, Laurent SE, Richardson J, Mitchell BC, Meier HCS. The legacy of structural racism: associations between historic redlining, current mortgage lending, and health. SSM Popul Health. 2021;14 doi: 10.1016/j.ssmph.2021.100793. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0008] 8.Tram KH, Saeed S, Bradley C, et al. Deliberation, dissent, and distrust: understanding distinct drivers of coronavirus disease 2019 vaccine hesitancy in the United States. Clin Infect Dis. 2022;74(8):1429–1441. doi: 10.1093/cid/ciab633. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0009] 9.Kim D. Associations of race/ethnicity and socioeconomic factors with vaccination among US adults during the COVID-19 pandemic, January to March 2021. Prev Med Rep. 2023;31 doi: 10.1016/j.pmedr.2022.102021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0010] 10.Lindemer E, Choudhary M, Donadio G, Pawlowski C, Soundararajan V. Counties with lower insurance coverage are associated with both slower vaccine roll out and higher COVID-19 incidence across the United States. Vaccines (Basel) 2021;9(9):973. doi: 10.1101/2021.03.24.21254270. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0011] 11.El-Mohandes A, White TM, Wyka K, et al. COVID-19 vaccine acceptance among adults in four major US metropolitan areas and nationwide. Sci Rep. 2021;11(1):21844. doi: 10.1038/s41598-021-00794-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0012] 12.Skolnik A, Bhatti A, Larson A, Mitrovich R. Silent consequences of COVID-19: why it's critical to recover routine vaccination rates through equitable vaccine policies and practices. Ann Fam Med. 2021;19(6):527–531. doi: 10.1370/afm.2730. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0013] 13.Kipnis P, Soltesz L, Escobar GJ, Myers L, Liu VX. Evaluation of vaccination strategies to compare efficient and equitable vaccine allocation by race and ethnicity across time. JAMA Health Forum. 2021;2(8) doi: 10.1001/jamahealthforum.2021.2095. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0014] 14.Maizlish N, Delaney T, Dowling H, et al. California healthy places index: frames matter. Public Health Rep. 2019;134(4):354–362. doi: 10.1177/0033354919849882. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0015] 15.King NB, Harper S, Young ME. Use of relative and absolute effect measures in reporting health inequalities: structured review. BMJ. 2012;345(7878):e5774. doi: 10.1136/bmj.e5774. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0016] 16.Katuramu R, Wallenta J, Semitala FC, et al. Closing the gap: a novel metric of change in performance. East Afr J Appl Health Monitor Eval. 2019;2019(3) http://eajahme.com/wp-content/uploads/2019/02/Katuramu_FINAL.pdf [PMC free article] [PubMed] [Google Scholar]

[bib0017] 17.Blueprint for a safer economy. California Department of Public Health. https://www.cdph.ca.gov/Programs/CID/DCDC/Pages/COVID-19/COVID19CountyMonitoringOverview.aspx. Updated June 15, 2021. Accessed November 17, 2022.

[bib0018] 18.Vaccination data. CA.gov. https://covid19.ca.gov/vaccination-progress-data/#equitably-across-groups. Updated February 16, 2023. Accessed November 22, 2021.

[bib0019] 19.2015-2019 ACS 5-year Estimates. U.S. Census Bureau. https://www.census.gov/programs-surveys/acs/technical-documentation/table-and-geography-changes/2019/5-year.html. Updated December 8, 2021. Accessed October 22, 2021.

[bib0020] 20.As California surpasses 30 million vaccines, Governor Newsom doubles down on efforts to vaccinate hard-to-reach communities. Sacramento, CA: CA.gov. https://www.gov.ca.gov/2021/05/04/as-california-surpasses-30-million-vaccines-governor-newsom-doubles-down-on-efforts-to-vaccinate-hard-to-reach-commun. Published May 4, 2021. Accessed January 21, 2022.

[bib0021] 21.Barry V, Dasgupta S, Weller DL, et al. Patterns in COVID-19 vaccination coverage, by social vulnerability and urbanicity — United States, December 14, 2020–May 1, 2021. MMWR Morb Mortal Wkly Rep. 2021;70(22):818–824. doi: 10.15585/mmwr.mm7022e1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0022] 22.Batteux E, Mills F, Jones LF, Symons C, Weston D. The effectiveness of interventions for increasing COVID-19 vaccine uptake: a systematic review. Vaccines (Basel) 2022;10(3):386. doi: 10.3390/vaccines10030386. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0023] 23.Braveman P. Health disparities and health equity: concepts and measurement. Annu Rev Public Health. 2006;27:167–194. doi: 10.1146/annurev.publhealth.27.021405.102103. [DOI] [PubMed] [Google Scholar]

PERMALINK

Use of the Improvement Index to Evaluate Equitable COVID-19 Vaccine Allocation in the San Francisco Bay Area

Riana B Jumamil, MD

George Rutherford, MD

HIGHLIGHTS