Similar to how Salesforce famously declared an “end to software” 20 years ago, today we say goodbye to ZIP codes as legacy tools in market data analysis for healthcare.
While Salesforce didn’t exactly end the use of software – it just moved software to the more accessible cloud – it revolutionized how companies conducted their business. Spatially Health, like the Salesforce of 2000, has taken a novel approach to understanding what influences individuals’ healthcare needs and their decisions to seek care. Missing from the approach – ZIP codes. Like software of the past, ZIP codes have gone to the proverbial graveyard.
Use of ZIP Codes in Marketing, Healthcare Reimbursement
Marketing and sales professionals like ZIP codes, because they can be used to target a certain geographic location. They even lend insight into some characteristics of people living within those ZIP codes. These characteristics include some demographic information – age, sex, household income – and some previous buying behavior. These geographic boundaries work to simplify and summarize a potential target audience and their consumption of goods and services within a specific area.
Additionally, the Centers for Medicare and Medicaid Services, or CMS, uses ZIP codes to determine reimbursements for medical services.
With few other tools available to provide a snapshot of potential patients and plan members, physicians, payers, healthcare executives and strategists followed the lead of marketers and CMS and for years have been relying on ZIP codes to develop business strategies, plan and build provider networks.
Pitfalls of ZIP Codes in Healthcare Analytics
Using ZIP codes to build healthcare business strategies and account for needs falls short, though. That’s because the way individuals seek care and use healthcare services differs greatly from buying clothing or household goods.
Knowing individuals’ age, sex, household income and what they purchase online over the past year cannot reveal why they may seek care from a certain doctor or hospital, or why they may choose one health plan over another. For that granular level of insight, deeper analysis is required. With that deeper analysis comes the need for more data points to consider. And just like the software that Salesforce denounced for its inability to quickly adapt to changing technology and behaviors, stagnant or incomplete data doesn’t help to reveal or predict healthcare’s local trends and patterns. The healthcare industry no longer needs to accept ZIP codes as the tool to reach their target audiences.
Spatial Intelligence to Predict Healthcare Behavior
People don’t equate with ZIP codes. Where a person lives, works and plays, or his physical context is much more meaningful than the arbitrary boundary that was devised by the government to make mail delivery – not healthcare delivery – easier. There is way too much variation in people’s behaviors and their connections to their surroundings that a ZIP code doesn’t factor in. In fact, individuals may have more similarities with those living, working and recreating across the imaginary ZIP code boundary, just a few blocks from them, than they do with those living within their designated ZIP code, which can span several dozen miles, depending on the population density of that area.
Research, such as that of epidemiologists to track where disease outbreaks originate, doesn’t rely on ZIP codes. Actuaries and financial analysts don’t focus on ZIP codes to predict their outcomes and advice. That’s because these groups have figured out the factors they’re studying don’t fit neatly into man-made boundaries. Neither does healthcare usage. We are looking at real problems but are relying on the wrong tools to help us address them.
Context Is Key to Healthcare Analytics
Consider the number of primary care practices in a ZIP code with 34,750 people. Looking at the example map below, which outlines the ZIP code boundaries of a 6.1 square-mile geographic area, it would seem that one primary care practice is expected to handle the demand of all 34,750 people living in that ZIP code. But when you look outside the boundaries of the ZIP code, you see that those 34,750 people actually have access to seven primary care practices that are fairly well distributed. Context is key in this example. The spatial relationship between where people live and where they may seek care cannot be expressed with arbitrary boundaries such as ZIP codes.
Locations of primary care physicians in and around a ZIP code.
ZIP Codes Lack Uniformity
Similarly, the following map illustrates an often ignored truth about ZIP codes — people aren’t distributed uniformly across them, and where people live strongly influences their healthcare needs and usage patterns. In this example, higher concentrations of people live in the southern and northern parts of the area. So, any healthcare planners and strategists would be better off looking at these concentrations across the map than only within the arbitrary boundaries of the ZIP code.
Distribution of people within a ZIP code.
ZIP Codes Skew Data
For payers, who wish to develop provider networks that will support members’ needs, ZIP codes may mistakenly reveal that an area is more affluent than it really is, simply because the median house value within a ZIP code may be higher than a neighboring one. In the map below, there are sections of the ZIP code that are indeed affluent, but the values of the overall area may be skewed upwards due to the fewer number of homes that have higher values. Location intelligence, as represented by the map, accounts for these differences and shows the value of applying these principles, rather than just relying on ZIP code boundaries.
Just because a ZIP code’s median home value is high, doesn’t necessarily mean that it’s an affluent area.
ZIP Codes Can’t Predict Healthcare Needs
Finally, the example map below shows how the government configures ZIP codes. It essentially measures the mail volumes of residents in certain areas and creates an arbitrary boundary around geographic areas that meet its threshold for those volumes, calling it a ZIP code. For healthcare, this type of grouping has no benefit, as the only similarities among those residents is that they live somewhat in close proximity to each other. But, the similarities, especially about their healthcare needs, usage patterns, and their healthcare decisions end at the imaginary boundaries.
These maps show ZIP codes and reveal how arbitrary they truly are.
Spatial Intelligence to Replace ZIP Codes for Healthcare Data Analysis
Since using what is measured within a ZIP code doesn’t allow us to consistently analyze the necessary data to predict healthcare needs and usage patterns, we need a different unit of measurement. This is when we turn to spatial intelligence, which gives us context to a person’s choices.
Spatial intelligence has existed and been used successfully by other industries. Now, Spatially Health is applying its principles and methodologies to healthcare to help our clients see the larger picture of what factors influence healthcare usage. We’ve developed a proprietary tool that includes the demographics that using ZIP codes attempts to capture, but with the geographic integrity and privacy required by HIPAA. And since we’ve developed this tool in a way that privacy doesn’t have to be a casualty of advanced analytics or stifle innovation, it allows us to form predictive models that help healthcare strategists and planners build future services and provider networks that meet the needs of their current and prospective patients and plan members. Find out how we can apply this tool to your needs.
Death of ZIP Codes for Healthcare Analytics
Much like the 21st Century claim by Salesforce that software was soon to be replaced by a more accessible and flexible application of the tool, Spatially Health has taken the legacy of using ZIP codes to categorize and plan healthcare services for individuals living within those boundaries and built a better tool. Spatially Health’s innovative approach uncovers people’s connections to their surroundings and their interactions in a uniform, consistent and agile manner that is effective for analytical purposes and conducive to bringing positive impacts and health outcomes to patients and members.
Rest in peace ZIP codes.