The COVID-19 Vaccine Dispensing Plan Disenfranchises At-need Populations

*Reposted from my Linkedin Feed*

I did a study a while back, building on a research project completed by my cohort at the Los Angeles Emergency Management Department (LAEMD) on the vulnerability of supply chains in Los Angeles, post-earthquake.

We discovered that at-need population centers had limited access and availability to commercial pharmacies like CVS and Walgreens.

The current plan for initial COVID-19 Vaccination distribution is through these big-box pharmacies. As the at-need population centers have pharmacy deserts, those affected locations may have limited access to the vaccine.

These groups tend to have the highest Covid-19 infection, hospitalization, and fatalities over any other grouping. This is why it is important to understand the difficulty of reaching these populations by relying on dispensing from a large retail pharmacy box store and the ontology of pharmacies.

Below is my abstract and GIS Story Map for studying a defined at-need population using specific selected factors, my methodology, and the pharmaceutical supply chain's availability after a large earthquake in Los Angeles.

I propose identifying the City of Los Angeles' at-need population areas and their ability to access pharmaceutical goods versus an established model by using my determined essential correlating indicators for an at-need population for Los Angeles. This study will also locate "pharmaceutical deserts" that may exist within the most vulnerable communities with the City of Los Angeles and ensure their resiliency in the event of a disaster. Los Angeles, a city with four million, is at risk for a catastrophic event in a massive earthquake accompanied by wide-spread devastation. Per the United States Geological Survey's (USGS) third Uniform California Earthquake Rupture Forecast (UCERF3) of 2015, a 7.5 or greater earthquake spawned by a shift in the San Andreas fault has a 16.2 percent probability of occurring in the next 30 years. Considering the potential damage and devastation an event like this could create, the supply of pharmaceutical goods must continue to flow into affected areas post-disaster: specifically, the sites that contain sizeable at-need density populations. The Northridge Earthquake of 1994 exposed critical issues of risk readiness and supply chain resilience for the City of Los Angeles. The earthquake hazard and the potential damage to the pharmaceutical supply chain for the City have been identified. Now the question becomes, how does one correctly identify vulnerable populations within the City of Los Angeles, and by what metrics?

Out of 16 hazards identified by the Federal Emergency Management Agency (FEMA), the City of Los Angeles is prone to 13 of them, including earthquakes. That identified hazard can disrupt or completely stop normal traffic flow and affect the transportation of goods into Los Angeles. For this study, an earthquake model was used to highlight the dangers of pharmaceutical supply disruption for the City of Los Angeles. I used the 2008 United States Geological Survey (USGS) "The Great California ShakeOut – Southern San Andreas Shakeout" earthquake exercise as the underlying layer of my study. USGS based their KML shakemap from their shakeout exercise on a scenario involving a magnitude 7.8 earthquake along the southernmost San Andreas Fault, originating from the Salton Sea.

 Currently, the County of Los Angeles relies solely on resources available through the Strategic National Stockpile and similar programs to provide pharmaceuticals to the City's residents during a disaster, but that only addresses bioterrorism and common outbreaks of the flu virus. In a disaster, these caches of medications are not suited or designed to replace the prescriptions of those if an earthquake with chronic conditions suddenly disrupts the pharmaceutical supply chain. Therefore, it is direly important to identify the vulnerable areas within Los Angeles.

Distribution Centers (DC) are an essential conduit between the pharmaceutical manufacturer, who creates and manufactures the drug, and the demand nodes, such as hospitals and pharmacies, to whom the drug is delivered. During a disaster, the pathway between distribution centers and demand nodes must remain secured. There are three pharmaceutical distribution centers located within the boundaries of Los Angeles County that serve the City of Los Angeles. Two are co-located in Valencia, east of the I-5 and northeast of the I-126, while the third resides in Santa Fe Springs, east of the I-605 and the Rio Hondo Channel, while West of Route 72. According to the 2008 USGS earthquake Shakeout Earthquake Scenario, all three are located in a heavy movement zone, which means all three are at substantial risk of damage if a large magnitude earthquake were to occur. Lives will be saved if these supply chains are resilient. However, if the pharmaceutical supply chain fails, tens-of-thousands of people will be at serious risk of a health crisis.

In the State of California, the five most prevalent chronic conditions include high blood pressure, diabetes, severe psychological distress, asthma, and heart disease. Per a 2011-2012 UCLA California Health Interview Survey, 28.5% of adults have one or more chronic conditions, which has decreased from a similar study conducted in 2007. In the 2007 survey, 34.8% of adults in Los Angeles County had one or more chronic conditions. South Los Angeles had the most adults with one or more chronic conditions at 38.5%, with high blood pressure as the most prevalent chronic condition. The other areas had significantly fewer adults with one or more chronic conditions: San Fernando (35.3%), East area (36.2%), West (34.3%), South Bay (35.1%), and Metro (30.6). South Los Angeles has the highest rate of its population with one or more chronic conditions.

 Pharmaceuticals encompass a wide array of illnesses, and this study acknowledges the criticality of identifying these population centers to ensure they are the priority for resupply post-disaster. This study seeks to ensure residents with chronic conditions, though ill, can maintain a homeostatic state supported by their prescribed and uninterrupted drug regimen. This is for residents whose pharmaceuticals for their chronic conditions can independently be taken at home (i.e., pills). The overall goal is to bolster and ensure the pharmaceutical supply chain is resilient enough to supply the necessary medications to the City of Los Angeles residents and prevent chronic conditions from developing an acute infection. Symptoms worsen quickly due to a lack of prescribed medications. When a person with a chronic disease from an illness previously stabilized by now absent medications due to a catastrophic event may place a heavy resource burden on first responders and hospitals, they provide care to acutely injured victims. The Red Cross - Los Angeles Region in 2011 sought to identify vulnerable communities within The City of Los Angeles and published their research in a study called "PrepareLA." In this study, eight socioeconomic and demographic factors were used to identify vulnerable communities within the City. These eight factors include (1) race/ethnicity and poverty, (2) limited English proficiency, (3) single-parent head of households, (4) educational attainment, (5) car-less households, (6) age dependency ratio, (7) population index, and (8) the presence of individuals with access and functional needs. With those factors, the study identified ten areas of vulnerability in Los Angeles. These areas include (1) Westlake, (2) Historic South Central, (3) South Park, (4) Central Alameda, (5) Pico-Union, (6) Florence, (7) Watts, (8) Cudahy, (9) Boyle Heights, and (10) Koreatown. As I continued my research, I felt some of the factors used to identify vulnerable populations centers did not directly point to vulnerability, rather causation from the other factors.

An example would be the factor of "car-less households" Does the fact of not owning a vehicle correlate with being vulnerable or caused by the element of poverty? I also felt that other factors like the "population index" unfairly focused on the vulnerable community outcomes in favor of the City of Los Angeles. In contrast, other communities outside the City have fewer resources. They are less likely to receive prioritized help in the first 72 hours if there is a dramatic disruption in the pharmaceutical supply chain. I chose factors I felt better represented the correlations between being a vulnerable resident and identifying a vulnerable person. In preparing and planning for a disaster, the City of Los Angeles' residents with chronic conditions, particularly regarding medication, must be identified by apparent correlating factors.

The resources and datasets for my project were from a variety of locations. The point data was self-created by me as part of the supply chain resiliency study for the City of Los Angeles. The data points reflect the LA City and County pharmacies and LA City hospital pharmacies locations as they currently exist. Some of the LA County supply nodes for these pharmacies lie outside the LA County boundary. The Boundary data was from the Los Angeles GeoHub, an open-source portal for GIS data. As identified from the 2011 Red Cross study, the vulnerable areas were drawn by me as described by the report at street level and intersections. The Tiger data and the Poverty Census data came from Census.gov.

I used the primary key "GEOID" to join LA County Tiger data with the Poverty Census Data tables, both obtained from Census.gov so that I could georeference the attribute data needed for my factors. With my identified factors of residents designated to be living in poverty (HC02_EST_VC05), residents that have not completed their High school education (HC01_EST_VC28), and residents 18-64 years of age that are unemployed (HC01_EST_VC39), I captured the mean from the Statistics report in Arcmap. From that, I could select tracts where people designated a being in poverty numbered more than 500, people who had not completed their high school education numbered more than 600. People in an unemployed status numbered more than 30.

I then used the ArcMap field calculator to run a SQL query on my selected factors, as seen below.

SQL Statement

Select *

From LA_County_Joined_Income

Where HC02_EST_VC05>500 and HC01_EST_VC28 > 600 and HC01_EST_VC39>30

 I combined the normalized pharmacy data and added the pharmacy layer. I then clipped everything to the City of Los Angeles Boundary. I then performed a join selection by location and joined the pharmacies to the SQL query results. I buffered the pharmacies with a .3-mile buffer. I then completed another join with the vulnerable areas to the joined pharmacy/vulnerable population as indicated with my factors and displayed my findings in a GIS Storymap format. See above.

The results presented as I predicted. Outside of the City of Los Angeles, many areas reflect and fit my vulnerable community factors. My study shows that population density skewed the previous study results to focus only on high population density areas, namely the City of Los Angeles. Without that factor, my results show large pockets of vulnerability in the rural areas, where medication resupply and priority to resupply are below the City of Los Angeles.

I joined the pharmacy layer to the previously identified vulnerable layers, along with my factor results. The results showed that my vulnerable area factors only occupied, at most, half the tracts and in some areas, none of the regions that used eight factors for vulnerable area designation.

Unusually enough, there was a lack of pharmacies that were not necessarily in the category of being vulnerable, as listed by my factors in some areas (unemployment, poverty, education). For example, Central Alameda showed they did not meet my at-risk area requirements but lacked any pharmacy or hospitals. Compare that to Watts, where more than half of the tracts met my parameters for an at-need population but only have one pharmacy allocated for the area. Even though Koreatown has areas that show at-need locations, they have nine pharmacies and a single hospital located directly within their boundary. They are also located in a severe shake zone in the event of an earthquake.

What this showed, first and foremost, is that the identified population with the most hospitals and pharmaceutical resources are at the most significant risk from earthquake damage, per their location within the areas of the greatest shake potential. That could mean that residents in those at-need areas must travel outside their zone for their pharmaceutical needs. Due to a lack of resources in their regions, this self-migration may overload the hospitals and pharmacies with an added surge of unanticipated patients. It also revealed significant findings for both pharmaceutical wholesalers and retail companies. McKesson, AmerisourceBergen (ABC), and Cardinal Health represent the top pharmaceutical wholesalers, while CVS, Walgreens, and Rite Aid represent the leading retail drugstores. The wholesalers' distribution centers are located outside of the City of Los Angeles but still reside within the Southern California region; they are in Valencia and Santa Fe Springs. With the possibility of a large magnitude earthquake, these distribution centers are at risk of becoming severely damaged or non-operational. Within vulnerable communities, independent pharmacies outnumber larger pharmacies, including chain drugstores (i.e., CVS, Walgreens), grocery stores with pharmacies (i.e., Ralphs, Albertsons), and mass merchants with pharmacies (i.e., Target, Walmart).

I hope somebody will organize these analyses to provide a substantive background needed to grow awareness of the vulnerability of areas with the greatest resource while understanding an at-need vulnerable population that must be appropriately realized and attended to with the same priority as a more populated place. Ultimately, the goal is to obtain a sufficient understanding of typical pharmaceutical supply chain operations to ensure that resilience, restoration, and redirection are as effective as possible during an abnormal to the disastrous event. As a note, Red Cross LA has developed a new metric for identifying vulnerable populations for their region I am not yet privy. Also, my study only shows pharmacies located within the eight-factor identified vulnerable population areas.