You are writing a grant proposal for a cancer screening or patient navigation program. One source says your county has a high cancer rate. Another puts it near the middle. A third reports a percentage instead of a rate per 100,000.
Before choosing a number, check what each source is counting. Cancer incidence, prevalence, and mortality answer different questions. A useful needs statement makes that distinction clear and connects the evidence to the service you are proposing.
For county cancer incidence and mortality, a practical starting point is State Cancer Profiles, a joint National Cancer Institute and CDC resource. It provides geographic comparisons and information about cancer screening and risk factors. For modeled estimates of adults reporting a cancer diagnosis, consult CDC PLACES. Your state cancer registry may offer additional local detail.
The right measure depends on the question:
| Question in your proposal | Measure to examine | What to check |
|---|---|---|
| How many new cancers are being diagnosed? | Registry incidence and average annual case counts | Cancer type, reporting period, population, and age adjustment |
| How many people are living with a cancer history? | Cancer prevalence | Whether the estimate comes from a registry or survey, and which diagnoses and ages it includes |
| How many residents are dying from cancer? | Cancer mortality | Cause of death, reporting period, and whether the rate is crude or age-adjusted |
| Is there evidence of a screening or follow-up gap? | Screening participation, stage at diagnosis, and local follow-up records | Whether the measures cover the people and service your program would reach |
NCI defines incidence as newly diagnosed cases during a specified period, mortality as deaths during a period, and prevalence as people alive with a prior or new diagnosis at a specified time. These measures describe different parts of a community's cancer burden. NCI definitions
PLACES adds another distinction: its cancer measure is modeled from survey responses about a previous diagnosis. It is not a registry count of new cancers. Check the definition for the release you use, including which skin cancers are included or excluded. A percentage of adults reporting a cancer history should not be compared directly with an annual incidence rate for residents of all ages. CDC measure definition
Age adjustment is the next detail to check. An age-adjusted rate uses a standard population to reduce the effect of differences in age composition. When comparing a county with its state, use rates calculated with the same standard population, along with matching cancer definitions and reporting periods. NCI explanation of age-adjusted rates
That standardized rate is useful for comparison, but it is not a direct count of residents. Multiplying an age-adjusted rate by the county population will not recover the actual number of diagnoses. If you need to describe workload or the scale of services, look for the registry's average annual case count. For a screening program, you will also need an estimate of the eligible population; diagnosed cases alone do not tell you how many people need screening.
Here is a hypothetical example. Suppose a registry reports a county colorectal cancer incidence rate of 48 per 100,000 and a state rate of 40, both age-adjusted and covering the same five years. The county's point estimate is 20% higher. That calculation does not establish that the difference is statistically significant, and it does not explain why the rates differ.
A needs statement could say:
During [reporting period], [County] had an age-adjusted colorectal cancer incidence rate of 48 per 100,000 residents, compared with 40 statewide, according to [registry and table]. The county point estimate was 20% higher. We will interpret this comparison alongside the reported confidence intervals, local screening participation, and follow-up data when defining the population our proposed navigation program will serve.
The figures above are illustrative. Replace them with verified values and include the source link, reporting years, and access date. If you have evidence of missed appointments, incomplete referrals, transportation barriers, or delayed follow-up, describe that separately. It helps explain why your proposed service addresses a documented need.
A particularly tempting comparison is lower incidence alongside higher mortality. It can raise a reasonable question about delayed diagnosis. It cannot answer that question by itself. The two datasets may cover different periods and populations, and the people who died during one period are not necessarily those diagnosed during that period. A crude mortality rate also retains age differences that an age-adjusted incidence rate is designed to reduce.
For a proposal focused on earlier diagnosis, seek more direct evidence: stage at diagnosis where available, screening completion, time to diagnostic follow-up, or local accounts of barriers to care. Describe what the evidence supports and what remains uncertain.
Environmental indicators require the same care. A county may have elevated cancer incidence and documented industrial releases or pesticide use. Their presence in the same county gives you a question to investigate; it does not establish that those sources caused the reported cancers. A defensible proposal identifies the specific concern and the additional evidence needed to understand it.
In Banana Analytics, county profiles bring cancer incidence, mortality, and modeled health measures into the same place. The underlying indicators remain essential: a summary score cannot substitute for the cancer-specific measure your proposal needs. Use the county directory to begin exploring, and consult the methodology before interpreting scores.
If you are assembling the rest of your needs statement, our county health data sourcing guide covers the broader source selection process. For a small county with missing or suppressed values, see our guide to small and rural county health data.
A reviewer should be able to follow the path from your source to your comparison to your proposed service. Give them a clearly defined measure, a fair benchmark, and local evidence explaining who needs help. That is a stronger foundation than a dramatic ranking whose meaning is difficult to defend.