Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Data, Analytics & AI · Episheet

Episheet: Kenneth Rothman’s Spreadsheet Tool for Epidemiologic Calculations and Confidence Intervals

Episheet is not generic spreadsheet software but a specific, author-distributed workbook developed by the epidemiologist Kenneth Rothman. It computes risk ratios, odds ratios, standardised measures and their confidence intervals using the formulas set out in Modern Epidemiology. This article describes what the tool computes and where it fits in epidemiologic analysis.

Pubrica AcademyWritten by a biostatistics reviewer Reviewed byan epidemiology methodologist Published10 October 2026 Last checked10 October 2026 Reading time8 minutes Sources10 peer-reviewed

Episheet is a free spreadsheet workbook, distributed by its author Kenneth J. Rothman, designed specifically to perform the manual epidemiologic calculations required for a 2x2 table, a standardized ratio, or a stratified Mantel-Haenszel analysis: risk ratios, rate ratios, odds ratios, and their confidence intervals, computed using the exact formulas documented in Rothman, Greenland, and Lash’s textbook Modern Epidemiology. Many of these calculations, although conceptually simple, involve formulas that are tedious and error-prone to compute by hand; presenting them transparently in a spreadsheet, rather than within the internal functions of statistical software, allows the user to inspect and verify every step.

Learning objectives

  • What Episheet computes, and why it is tied specifically to Rothman’s published formulas
  • The difference between the exact (Fisher-type) and approximate confidence interval methods it offers
  • Where Episheet fits relative to full statistical software packages such as Stata, SAS and R
  • How to validate that a result computed in Episheet is correct before relying on it

Common misconceptions

  • “Episheet is a general-purpose data analysis spreadsheet comparable to any templated Excel workbook.” Episheet is a specific, author-attributed tool built around the formulas set out in Rothman, Greenland, and Lash’s Modern Epidemiology. It is used to compute defined epidemiologic measures (risk ratios, odds ratios, standardized ratios, and Mantel-Haenszel pooled estimates) with their exact or approximate confidence intervals, and it is not a general template of charts and formulas interchangeable with any other spreadsheet.

Measures computed by Episheet

Episheet takes the counts from a 2x2 (or stratified multi-table) exposure-outcome layout and computes the standard epidemiologic measures derived from them: risk ratios and risk differences for cohort data, odds ratios for case-control data, and rate ratios where person-time data are available, together with confidence intervals for each1. For stratified data, it computes Mantel-Haenszel pooled estimates across strata, allowing the user to assess effect-measure modification between strata before deciding whether pooling is appropriate2. It also supports standardization, computing standardized mortality or morbidity ratios and directly or indirectly standardized rates; this calculation is straightforward in principle but error-prone by hand when several age strata and their corresponding population weights are involved2.

Exact versus approximate confidence interval methods

A distinctive feature of Episheet is that it typically offers more than one method for computing a confidence interval around the same point estimate, because different approximation methods perform differently depending on sample size and on how extreme the observed counts are2. Exact methods, based on the actual (often Poisson or binomial) distribution of the underlying counts, are generally preferred for small samples or rare outcomes, where approximate formulas can produce misleadingly narrow or asymmetric intervals; large-sample approximate methods, such as test-based or Wald-type formulas, are computationally quick and often adequate once the count in every cell is reasonably large3. The availability of both methods in the same worksheet allows the user to compare how sensitive a given finding is to the choice of method, which is itself a useful check when a reported interval appears unexpectedly narrow or wide for the sample size involved.

Position of Episheet relative to full statistical software

Episheet is not a substitute for full statistical software such as Stata, SAS, or R when an analysis requires multivariable regression, complex survey weighting, or the handling of large individual-level datasets; it is designed for tabulated counts and a relatively small number of strata, the type of summary data that commonly appears in the results table of a published paper rather than raw patient-level records1. Its appropriate uses are therefore reanalyzing or verifying summary data reported in a published table, teaching the manual logic underlying a risk ratio or odds ratio calculation, and computing a confidence interval for a result where the original paper did not report one; in each of these situations, using full statistical software to analyze a small number of cell counts would be disproportionate.

Validation of results before use

Because Episheet directly implements formulas documented in Modern Epidemiology, comparing its output with the textbook’s worked examples is a reasonable way to confirm that data have been entered correctly and that the correct output cell is being read, since the layout of the workbook can be unfamiliar on first use2. A second, independent check before a result is included in a manuscript is to recompute the simplest measure from the same 2x2 table by hand (the unadjusted risk ratio or odds ratio) and to confirm that it matches the unadjusted output of the workbook. A discrepancy very likely indicates that the counts were entered into the wrong cells, a common error because many epidemiologic 2x2 layouts place exposure and outcome on different axes, depending on the convention of the source textbook or software.

Appropriate use and limitations of Episheet

Users should also remain aware of a structural limitation common to tools of this kind: because Episheet works from already-tabulated counts, it inherits any confounding or bias present in the way those counts were originally stratified, and no confidence interval formula, exact or approximate, can correct for a confounder on which the data were never stratified3. The emphasis of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement on reporting how confounders were handled in observational studies is relevant here: a transparent account of which variables were stratified or adjusted for, whether the calculation was performed in Episheet or in full statistical software, enables a reader to judge whether the analysis addressed the confounding relevant to the specific research question4. Applied within that scope, namely aggregated or tabulated epidemiologic data requiring a transparent and verifiable calculation, Episheet remains a useful, if narrowly scoped, tool decades after its first release.

Continuing value for teaching and peer review

The continued use of Episheet decades after its original release is largely attributable to its transparency: because every formula is visible and editable in an ordinary spreadsheet cell rather than concealed within compiled statistical software, a student or peer reviewer can trace, step by step, how a reported risk ratio or confidence interval was derived from the underlying counts, rather than relying on an opaque function. This makes it an effective teaching tool for introducing the manual logic of standard epidemiologic measures before students progress to full statistical packages, which compute the same quantities automatically but do not display the intermediate arithmetic. Similarly, a peer reviewer who suspects that a reported confidence interval used an approximation method inappropriate for the sample size can reconstruct the calculation in Episheet directly from the numbers reported in the manuscript’s results table, without access to the authors’ original dataset or software.

Sources of Episheet’s confidence interval formulas

The confidence interval methods available in Episheet are not generic statistical defaults; they derive from named, published formulas in the epidemiologic methods literature. The test-based confidence interval method commonly offered for ratio measures derives from Miettinen’s description of estimability and estimation in case-referent studies, which set out how to obtain a confidence interval directly from a ratio estimate and its associated chi-square statistic rather than from a separately computed standard error5. The Katz log-transformation method, another option commonly available for risk ratio confidence intervals, derives from the 1978 paper by Katz and colleagues on obtaining confidence intervals for the risk ratio in cohort studies, which showed that computing the interval on the log scale before back-transformation produces better-behaved intervals than computing it directly on the ratio scale6. For stratified Mantel-Haenszel estimates, the variance formula derives from the 1985 paper by Greenland and Robins on estimating a common effect parameter from sparse follow-up data, which addressed a recognized weakness of earlier variance formulas when individual strata contain small cell counts, a situation frequently encountered in stratified epidemiologic analyses7. Knowledge of these formulas by name has practical value: it allows a user to consult the worked examples in the original paper when a result appears unexpected, rather than treating the output of the workbook as unverifiable. It also strengthens the methods section of a manuscript: stating which confidence interval formula was used (Miettinen’s test-based method, the Katz log method, or an exact method), rather than simply writing “95% confidence intervals were calculated,” enables a reader to check the calculation against its named source, as the analyst could, and is a minor addition that meaningfully improves the statistical transparency of a paper. The same principle applies to any secondary user of a published 2x2 table: if the original paper did not report its confidence interval method, recomputing the interval independently in Episheet using a named, documented method and stating that method explicitly is more transparent than reporting a recalculated interval without specifying how it was obtained.

Key points

Key points

Episheet is a specific, formula-transparent tool for computing risk ratios, odds ratios, standardized measures, and their exact or approximate confidence intervals from tabulated epidemiologic counts, not a general spreadsheet template, and its value is clearest when working from already-aggregated data rather than raw, individual-level datasets.

Practical application

Practical application

  1. Episheet is suited to calculations from tabulated counts, not individual-level regression. Full statistical software is required once an analysis needs multivariable adjustment, survey weighting or raw patient-level data; Episheet is designed for summary 2x2 or stratified tables.
  2. For small samples, the exact and approximate confidence interval outputs should be compared. When cell counts are small or an outcome is rare, it should be determined whether the exact and approximate methods diverge meaningfully before a single interval is reported.
  3. The first calculation should be validated against a worked example from Modern Epidemiology. This confirms that the workbook’s inputs and outputs are being read correctly before it is used for a calculation intended for a manuscript.
  4. The simplest unadjusted measure should be recomputed by hand as a second check. A basic risk ratio or odds ratio computed by hand should match Episheet’s unadjusted output before more complex stratified or standardised outputs are relied upon.
  5. Standardisation and pooling cannot correct for a confounder on which the data were not stratified. The variables requiring stratification or standardisation should be determined before counts are entered, since the tool’s formulas cannot retrospectively correct for an omitted stratification variable.

Frequently asked questions

No. Episheet refers specifically to the workbook distributed by Kenneth Rothman, built around the formulas documented in Rothman, Greenland, and Lash’s Modern Epidemiology; a generic spreadsheet with charts and basic formulas is not the same tool, even if it is used for similar purposes.

Different confidence interval formulas (exact methods based on the underlying distribution versus large-sample approximations) perform differently depending on sample size and on how extreme the observed counts are; offering both allows the user to assess how sensitive the result is to that methodological choice.

It is designed for tabulated counts and strata rather than raw individual records; for analyses requiring multivariable regression or the handling of large individual-level datasets, dedicated statistical software such as R, Stata, or SAS is more appropriate.

No. It can standardize or pool across strata that the user defines, but it cannot correct for a confounding variable on which the data were never stratified; deciding which variables to control for remains the analyst’s responsibility before any calculation begins.

References

  1. Rothman KJ. Episheet: Spreadsheets for the analysis of epidemiologic data. Boston University School of Public Health (author-distributed software and documentation). read 10 October 2026. View Episheet resources
  2. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Lippincott Williams & Wilkins; 2008. read 10 October 2026. Find this edition
  3. Wells GA, Shea B, O'Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Ottawa Hospital Research Institute. read 10 October 2026. View NOS documentation
  4. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. read 10 October 2026. View on PMC
  5. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. European Journal of Epidemiology. 2010;25(9):603-605. read 10 October 2026. View on European Journal of Epidemiology
  6. Kyu HH, Bachman VF, Alexander LT, et al. Physical activity and risk of breast cancer, colon cancer, diabetes, ischemic heart disease, and ischemic stroke events: systematic review and dose-response meta-analysis for the Global Burden of Disease Study 2013. BMJ. 2016;354:i3857. read 10 October 2026. View on The BMJ
  7. Sterne JAC, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919. read 10 October 2026. View on The BMJ
  8. Miettinen OS. Estimability and estimation in case-referent studies. American Journal of Epidemiology. 1976;103(2):226-235. read 10 October 2026. View on American Journal of Epidemiology
  9. Katz D, Baptista J, Azen SP, Pike MC. Obtaining confidence intervals for the risk ratio in cohort studies. Biometrics. 1978;34(3):469-474. read 10 October 2026. View on Biometrics
  10. Greenland S, Robins JM. Estimation of a common effect parameter from sparse follow-up data. Biometrics. 1985;41(1):55-68. read 10 October 2026. View on Biometrics

Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 10). Episheet: Kenneth Rothman’s Spreadsheet Tool for Epidemiologic Calculations and Confidence Intervals. Pubrica. https://pubrica.com/academy/concepts-definitions/episheet-framework/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

Episheet Use-Case Checklist

Confirm Episheet is the right tool before reaching for it.

[ ] Data are already tabulated counts (2x2 or stratified), not raw individual-level records
[ ] Analysis needed: risk ratio / odds ratio / rate ratio / standardized ratio / Mantel-Haenszel pooled estimate (circle one)
[ ] Confidence interval method chosen deliberately (exact vs approximate) based on sample size and rarity of outcome
[ ] Stratification variables decided before entering counts, not after seeing results
[ ] First calculation validated against a worked example from Modern Epidemiology
[ ] Simplest unadjusted measure recomputed by hand as a cross-check

Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.

How Pubrica can help

Pubrica’s biostatistics team uses tools such as Episheet for transparent, verifiable calculations on tabulated epidemiologic data, and full statistical software where individual-level or multivariable analysis is required, matching the tool to the data.

Find something else

Search the Pubrica Academy

Access is free and requires no account. Enter a topic, or open the full library to filter by category, sub-topic or resource type.

Selecting appropriate tools for calculations on tabulated data

A scoping assessment is free and takes two business days. It provides a written assessment of what the work requires and which specialist would undertake it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.