Description
SEER Cancer Outcomes Study
A population-level survival study built on the Surveillance, Epidemiology and End Results programme, designed for the way oncology journals now review registry work. SEER is the most accessible cancer outcomes resource in the world, which is exactly why editors have become sharp about it: the same handful of methodological errors appear in submission after submission, and reviewers look for them first. This service is built to survive that reading. For the underlying method, see our guide to what the SEER database can and cannot answer.
When This Is the Right Study
SEER covers roughly half the US population through its population-based registries and captures incidence, tumour characteristics, stage, first-course treatment, and cause of death with long follow-up. It is the right instrument for survival by stage and histology, temporal trends in incidence or outcome, disparities by race, insurance or geography, and the natural history of uncommon tumours where no trial exists and no single institution has enough cases.
It is the wrong instrument when your question depends on things SEER does not record. There is no comorbidity index, no performance status, no detailed chemotherapy regimen, no recurrence, and no data on treatment given beyond first course. Treatment comparisons drawn from SEER alone are confounded by indication in ways that no statistical adjustment fully removes, and reviewers know it. If your question is really a treatment-effect question, we will tell you that SEER-Medicare or a different design is the honest answer before you commit.
How the Analysis Is Actually Done
- Competing risks handled, not ignored – In an older cancer cohort, death from other causes is not censoring, it is a competing event. Kaplan-Meier applied to cancer-specific death in that setting overestimates the event probability. We report cumulative incidence functions and model with Fine-Gray subdistribution hazards alongside cause-specific Cox, and we state plainly which question each one answers.
- Proportional hazards tested, not assumed – Schoenfeld residuals examined for every covariate, with time-varying coefficients, stratification or restricted mean survival time used where the assumption fails. A Cox model presented without this check is the single most common reviewer objection to registry survival papers.
- Stage harmonised across editions – AJCC staging has changed edition to edition across the SEER window, and the derived variables are not interchangeable. We reconcile Summary Stage and Derived AJCC across the study period explicitly, or restrict to a window with consistent coding, and document which.
- Immortal time bias designed out – Any exposure defined by something that happens after diagnosis (receiving surgery, completing radiation) creates guaranteed survival time before it can be classified. We use landmark analysis or time-dependent covariates rather than the naive comparison that inflates the treated group’s survival.
- Relative and net survival where appropriate – Ederer II expected-survival methods with US life tables, and period analysis where recent prognosis is the question rather than historical cohort survival.
- Multiple primaries handled by an explicit rule – First primary only, or all primaries with clustering accounted for; either is defensible, silently mixing them is not.
- Unknown stage treated as a finding – Cases with unknown stage differ systematically from staged cases. We report them, test whether excluding them changes the conclusion, and say so.
- Case definition checked against the classification, not the code – Registries store legacy codes that predate current WHO classifications, so a cohort built by selecting on a database field is not always the cohort the clinical definition describes. We verify the mapping before the cohort is frozen.
- Treatment comparisons matched honestly – Where a comparison is defensible, propensity score matching or inverse probability weighting with covariate balance diagnostics, plus an E-value quantifying how strong unmeasured confounding would need to be to explain the result away.
What You Receive
- Complete manuscript – Introduction, methods, results, discussion and limitations, written for your target journal
- Survival figures – Kaplan-Meier or cumulative incidence curves with numbers at risk, forest plots, and trend figures, all publication-quality vector output
- Full analysis code – R or Stata, commented and runnable, including the SEER*Stat case selection
- Completed STROBE checklist – See our STROBE tutorial for what reviewers check
- Analysis log – Registry grouping, case definition, every inclusion and exclusion, every derived variable
- One round of author-requested revisions
What Is Not Included
We do not ghost-write authorship. You set the research question, review the interpretation, and own the scientific claims. We do not submit for you, do not guarantee acceptance, and do not handle peer-review response within this package. SEER-Medicare linked claims analysis is a separate, larger engagement. If the survival difference you hoped for is not there, we write the paper that reports that.
When We Will Tell You Not to Buy This
If your cancer site and outcome have been published from SEER repeatedly, another descriptive survival paper will not clear a good journal. If your cohort has fewer than roughly a hundred events, multivariable Cox will not support the number of covariates you want. If the question genuinely requires comorbidity or recurrence data, SEER cannot answer it and we will say so rather than take the work. The Feasibility & Protocol Assessment is designed for exactly this check.
Data Access
We can work either way. If you already hold SEER Research Plus access, send your SEER*Stat extract or case listing and we work from it. If you do not, we hold access and can build the cohort ourselves, nothing for you to arrange. Tell us which at the start so the analysis plan reflects it.
Timeline
Standard delivery is 15 business days from an agreed analysis plan. Multi-registry trend decomposition and matched treatment comparisons may take longer; if yours will, we say so before you pay.
Common Questions
Cause-specific or overall survival? Usually both. SEER cause-of-death coding is imperfect, so we report overall survival as the robust primary and cancer-specific as the mechanistically interesting secondary, rather than choosing whichever looks better.
Can you compare two treatments? Sometimes, with matching and an explicit confounding-by-indication limitation. We will not present a SEER treatment comparison as though it were causal.
Which registry grouping will you use? Whichever the question requires. SEER 8 for the longest trends, SEER 17 or SEER 21, which SEER renamed from SEER 22 when Massachusetts left the grouping, for contemporary population coverage. The choice is stated and justified in the methods.
I have a rejected SEER manuscript. Can you fix it? Frequently, yes, and usually for less than a new study. Send the manuscript and the reviewer comments.
Extended Scope
SEER-Medicare linkage, multi-registry pooling, competing-risk nomogram development with external validation, and joinpoint trend modelling are quoted individually, typically from $3,500. See the Extended & Multi-Database Registry Study.
Have a Question First?
If you are not sure this is the right fit for your research question, ask before you buy. Send it below and a senior analyst replies within one business day. There is no obligation, and if a different product suits you better, or if the honest answer is that we cannot help, we will say so.
