35 ChatGPT Prompts for Epidemiologists: From Data to Public Health Impact

#ai #research #chatgpt #productivity

Epidemiologists investigate disease distribution, identify risk factors, design surveillance systems, and translate complex data into actionable public health guidance. These 35 ChatGPT prompts help you move from data to insight faster, communicate findings more effectively, and design stronger studies. Copy, paste, and adapt them to your work.

1. Study Design & Methods

I am designing an epidemiological study to investigate the association between [exposure, e.g., particulate matter air pollution] and [outcome, e.g., incident asthma] in [population, e.g., children ages 5-12 in urban areas]. Compare the strengths and limitations of a cohort study, case-control study, and cross-sectional design for this specific research question. Recommend the optimal design and justify your choice.

Help me calculate the required sample size for a [study design] investigating [exposure-outcome association]. The following parameters apply: expected prevalence/incidence [X], odds/risk ratio of interest [Y], desired power [80/90]%, alpha [0.05], and estimated loss to follow-up [Z]%. Show the formula and walk me through the calculation step by step.

I am designing a case-control study on [topic]. Help me develop: the case definition (with inclusion and exclusion criteria), the control selection strategy and matching variables, the exposure assessment method, potential sources of selection bias and how to minimize them, and a recall bias mitigation strategy for self-reported exposures.

Evaluate the validity of the following exposure assessment approach for [exposure type] in a prospective cohort study: [describe method, e.g., single baseline questionnaire / biomarker measurement / administrative data linkage]. Assess: measurement error type (random vs. systematic), likely direction of bias in effect estimates, and suggest a more valid approach if available.

I need to design a cluster-randomized trial evaluating [intervention, e.g., a community health worker program] in [setting]. Address: rationale for cluster vs. individual randomization, unit of randomization, intraclass correlation coefficient (ICC) implications for sample size, randomization procedure, and how to account for clustering in the analysis.

2. Epidemiological Analysis

I have a dataset from a cohort study with the following variables: [list key variables including outcome, exposure, covariates, follow-up time]. Walk me through the analysis plan: how to calculate crude and adjusted incidence rates, which confounders to include in multivariable regression, whether to use Cox regression or logistic regression and why, how to handle time-varying exposures, and what sensitivity analyses to run.

Explain how to identify and handle confounding in my analysis of [exposure] and [outcome]. My dataset includes: [list potential confounders]. Describe: how to assess confounding using the 10% change-in-estimate rule, propensity score methods as an alternative, residual confounding risks, and how to present adjusted vs. unadjusted estimates in a table.

I am analyzing data from a cross-sectional survey of [N] participants. The outcome is [binary/continuous] and my key exposure is [variable]. I have the following covariates: [list]. Guide me through: appropriate regression model selection, how to handle missing data (complete case vs. multiple imputation), how to calculate prevalence ratios vs. odds ratios for a common outcome, and how to interpret the results.

Interpret the following epidemiological results for me: [paste results — OR/RR/HR with 95% CI and p-value]. Explain: the magnitude and direction of the association, statistical vs. clinical/public health significance, what the confidence interval tells us, and what I should say and NOT say about causality based on these results.

I want to conduct a meta-analysis on [topic]. Guide me through: defining inclusion/exclusion criteria for studies, the PRISMA process, how to extract and harmonize data across studies with different designs, how to select a fixed vs. random effects model, how to assess and quantify heterogeneity (I² statistic), and how to assess publication bias (funnel plot, Egger's test).

3. Disease Surveillance & Outbreak Investigation

Design a sentinel surveillance system for [disease/condition] in [setting, e.g., a national healthcare system / a low-resource country / a specific industry sector]. Include: objectives, case definition, data sources and collection points, reporting frequency, data flow architecture, analysis plan, and how surveillance findings will trigger public health action.

Walk me through the steps of a field outbreak investigation for a suspected foodborne illness cluster affecting [N] people at [setting, e.g., a wedding reception / a school / a correctional facility]. Follow the classic 10 steps of outbreak investigation, specifying what I should do in the first 24 hours, first week, and in the final report.

I am analyzing syndromic surveillance data and have detected an anomaly: [describe the signal, e.g., a 2.5 SD increase in ED visits for gastrointestinal illness in [city] over the past 3 days]. How do I: assess whether this is a true signal or noise, generate hypotheses about the cause, decide when to escalate to a full investigation, and communicate the preliminary signal to public health leadership?

Develop a line list template and attack rate table for an outbreak investigation of [disease/syndrome] at [setting]. Include: all variables to collect (demographic, clinical, exposure, laboratory), how to calculate attack rates by exposure category, and how to use the attack rate table to identify the most likely vehicle or source.

Write the Methods and Results sections of a preliminary outbreak report for [describe outbreak scenario: disease, setting, case count, dates]. Include: case definition, descriptive epidemiology (person/place/time), epi curve description and interpretation, attack rates, and preliminary hypothesis. Format appropriate for a health department situation report.

4. Public Health Communication & Translation

Translate the following epidemiological finding into a press release for a general public audience: [paste finding, e.g., adjusted OR 1.47 (95% CI 1.21-1.79) for [outcome] among [exposure group]]. The press release should: accurately represent the finding without overstating causality, contextualize the risk, provide actionable guidance, include a quote from a public health official, and avoid fear-inducing language.

Write a health advisory for clinicians about [emerging disease / outbreak / vaccine update]. Include: epidemiological summary (case counts, geographic spread, at-risk populations), clinical presentation and case definition, diagnostic and testing guidance, treatment/prophylaxis recommendations, reporting requirements, and key resources. Tone: authoritative, actionable, concise.

I need to present complex risk data to a [city council / hospital board / community group] that has no epidemiology background. The data shows: [describe finding, e.g., a 1 in 500 lifetime risk of [disease] associated with [exposure]]. Help me explain this risk in three different formats: absolute risk, relative risk framing, and a visual metaphor or natural frequency format that is most intuitive for lay audiences.

Draft a risk communication message for a [disease outbreak / environmental health threat / vaccine hesitancy situation] following the CERC (Crisis and Emergency Risk Communication) principles. Address: acknowledging uncertainty, what we know and don't know, what people should do right now, where to get trusted information, and how we will provide updates.

Write a policy brief (2 pages maximum) summarizing the evidence on [public health topic, e.g., the health effects of [exposure] / the effectiveness of [intervention]] for a state health department audience. Include: key findings from the literature, quality of evidence, public health significance, policy options and trade-offs, and a specific recommendation. Use clear subheadings and avoid statistical jargon.

5. Infectious Disease Epidemiology

Explain the key epidemiological parameters I need to characterize a novel infectious disease outbreak: R0 vs. Rt, serial interval, incubation period, generation time, case fatality rate vs. infection fatality rate, and ascertainment fraction. For each parameter, explain: what it measures, how to estimate it from early outbreak data, its limitations, and its implications for outbreak control.

Design a seroprevalence study to estimate the true burden of [infection] in [population/region] following an outbreak in which confirmed case counts are thought to underrepresent true infections. Cover: study design, target population and sampling strategy, sample size considerations, antibody assay selection and performance characteristics, analysis to adjust for test sensitivity/specificity, and ethical considerations.

I am modeling the impact of a vaccination campaign on [disease] transmission in a population with [vaccination coverage %] and vaccine efficacy of [X]% against infection and [Y]% against transmission. Help me calculate herd immunity threshold, effective reproduction number post-vaccination, and describe the key assumptions and limitations of this simple model.

Develop a contact tracing protocol for [disease] in [setting, e.g., a school / a workplace / a low-resource community setting]. Include: case interview procedure and exposure window, how to define a close contact for this pathogen, notification approach, quarantine vs. isolation guidance, monitoring of contacts, and data management and privacy considerations.

Summarize the current evidence on the effectiveness of [non-pharmaceutical intervention, e.g., masking / ventilation / school closures] in reducing transmission of [pathogen]. Describe: study designs used, effect size range, heterogeneity in evidence, key methodological limitations, and what the evidence does and does not support in terms of policy recommendations.

6. Chronic Disease & Environmental Epidemiology

I am investigating a potential disease cluster of [disease, e.g., childhood leukemia] near [exposure source, e.g., an industrial facility]. What is the appropriate epidemiological approach? Discuss: statistical methods for cluster detection and evaluation (SaTScan, standardized incidence ratios), the limitations of cluster investigations, how to communicate findings to a concerned community, and when a cluster warrants a full analytic study.

Design a study to evaluate the long-term health effects of occupational exposure to [chemical/agent] in [industry sector]. Choose the most appropriate study design, describe exposure assessment using job-exposure matrices or biomonitoring, identify the key health outcomes to track, address the healthy worker effect as a potential bias, and describe ethical considerations for worker research.

Summarize the Bradford Hill criteria for evaluating whether an observed association between [exposure] and [outcome] is likely causal. For each criterion (strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, analogy), apply it to the evidence for [specific exposure-outcome pair] and render a judgment on causality.

I have geocoded health outcome data and environmental exposure data for [geographic area]. Walk me through a spatial epidemiology analysis plan: how to map outcome rates, how to assess spatial autocorrelation (Moran's I), how to conduct a spatial regression to assess the exposure-outcome association while accounting for spatial autocorrelation, and how to visualize and present results.

Develop a life course epidemiology framework for studying the developmental origins of [chronic disease, e.g., cardiovascular disease / obesity / mental health disorders]. Include: key sensitive periods to study (prenatal, early childhood, adolescence), study design implications (birth cohort), relevant exposures at each life stage, and how epigenetic mechanisms might mediate these associations.

7. Grant Writing, Publishing & Professional Development

Help me write a Specific Aims page for an NIH R01 grant application studying [research question]. Include: an attention-grabbing opening paragraph on the public health significance, a clear statement of the knowledge gap, a brief overview of preliminary data, three specific aims with hypotheses, and a closing paragraph on innovation and impact. Length: 1 page.

Write an abstract for a manuscript reporting the results of a [study design] study examining [exposure-outcome association] in [population]. Include: background (2 sentences), methods (study design, population, exposure and outcome measures, analysis), results (key quantitative findings with effect estimates and confidence intervals), and conclusions (1-2 sentences, no overclaiming). Target journal: [journal name].

Review this Methods section of my epidemiology manuscript and provide feedback: [paste Methods]. Evaluate: completeness against STROBE/CONSORT/PRISMA reporting guidelines, clarity of the study design description, adequacy of the statistical methods section, whether I have described confounding control adequately, and any methodological concerns a reviewer is likely to raise.

I received the following peer review comments on my manuscript: [paste reviewer comments]. Help me draft a point-by-point response letter. For each comment: acknowledge the reviewer's concern, describe how I addressed it (or respectfully explain why I did not), and indicate where changes appear in the revised manuscript. Tone: collegial, thorough, and professionally assertive.

Design a 12-month professional development plan for an early-career epidemiologist wanting to establish a research program in [area, e.g., infectious disease / environmental health / health disparities]. Include: skills to develop (methods, communication, grant writing, leadership), conferences to attend, collaborations to build, a publication goal, and a target grant mechanism to apply for by year's end.