One way Pa. can help solve growing vaccine disparities between counties | Opinion

The Food and Drug Administration’s endorsement of Johnson & Johnson’s single-dose vaccine is another triumph in the fight against COVID-19. But as more vaccines are approved, still urgently needed is updated federal guidance on how Pennsylvania and other states should determine the number of doses counties need. Many counties are reporting acute vaccine shortages under current protocols and many of the most vulnerable — often in communities of color and rural areas — are being left behind.

Currently, the only official guidance on vaccine allocation comes from the CDC’s Advisory Committee on Immunization Practices (ACIP), whose recommendations have served as a vital starting point for states in terms of deciding how to allocate their limited supply of the vaccine. But ACIP’s model — which relies on a combination of population size and the Social Vulnerability Index (SVI), a tool designed to help government deliver aid in the aftermath of natural disasters, not necessarily pandemics — can be improved by incorporating real clinical data and social determinants of health (SDOH), such as air quality and access to fresh food.

Some officials have overlooked the vital role of county health departments in the vaccine distribution effort. The need for rigorous coordination and planning between states and the federal government is well understood, but the final step in distribution — administering vaccinations — often plays out at the county level. County health departments, 90% of which administer direct immunization services, are the gatekeepers of local health services, typically in charge of ordering the number of doses needed for residents who qualify for the vaccine.

» READ MORE: Pennsylvania admits vaccine undersupply in Philadelphia suburbs, promises to fix it with extra doses immediately | Maria Panaritis

Yet many states still lack an allocation model that ensures every single county receives enough doses. ACIP recommends allocation based on the size of priority populations in a given county, such as people with significant comorbid conditions. It also suggests reserving a certain percentage of doses for the most vulnerable communities based on SVI. While SVI factors in census data on some important variables that correspond to individual health, such as housing density and status of employment, it does not consider other critical risk factors associated with COVID-19, including for social determinants of health.

The result has been the same in many counties nationwide: unfilled orders, backlogged vaccination schedules, and mounting frustration with competition for highly sought-after doses.

My data science company Cogitativo just simulated the CDC’s ACIP’s approach for the 10 most populous states in the U.S. What we found, published in a report titled “No Relief In Sight,” was striking: 42% of counties in Pennsylvania will see a dose deficit under the ACIP approach, with a total combined shortage of 264,290 doses.

A more precise way to distribute the vaccine is to use SDOH and real clinical data about the prevalence of certain health conditions to determine the number of doses a county needs. These factors can help health officials make more informed decisions when allocating doses. For example, SDOH — conditions in the environments where people live that directly impact their health — can detect risks for those in underserved communities, such as proximity to local highways or access to groceries. At the same time, claims data from insurers can help highlight where there are higher concentrations of populations dealing with underlying health conditions — such as obesity or diabetes — that could make them more likely to suffer worse health outcomes if exposed to the virus and that need greater doses.

» READ MORE: Pa.’s ‘scavenger hunt’ approach to vaccines leaves everyone on their own | Opinion

Our analysis of the 10 most populous states found that such an approach could prevent 4.4 million more hospitalizations than would the ACIP model would. In Pennsylvania, specifically, the targeted approach would prevent roughly 268,500 more hospitalizations.

The U.S. passing the grim milestone of 500,000 deaths from the virus should serve as a sobering call to action for health officials that there is an urgent need to reach our most vulnerable populations and avoid many of these preventable deaths.

The most ambitious vaccination operation in our nation’s history was sure to present many challenges. But more effective coordination between states and counties can better support those who are too often overlooked and underserved. Moving toward a data-based approach that most effectively prioritizes vulnerable populations will ensure that no community in Pennsylvania is left behind.

Gary Velasquez is the cofounder and CEO of Cogitativo, a data science company.