Title 3: A Strategic Guide to Problem-Solution Framing and Avoiding Common Pitfalls

Every epidemiology project begins with a question. But how that question is framed—whether as a problem to solve or a solution to test—can determine the entire trajectory of the work. Get it right, and your investigation aligns resources, clarifies objectives, and leads to actionable insights. Get it wrong, and you risk wasted effort, misinterpreted data, or even harmful recommendations. This guide is for epidemiologists, public health researchers, and program managers who want to master problem-solution framing and sidestep the pitfalls that trip up even experienced teams.

1. The Decision Frame: Who Must Choose and by When

Framing a problem-solution structure starts with a decision: who is the audience, and what choice do they face? In epidemiology, the decision maker might be a local health department deciding whether to deploy a mobile clinic, a policy maker choosing between mask mandates and ventilation upgrades, or a research team selecting a study design. Without a clear decision frame, your framing will lack urgency and relevance.

Consider a typical scenario: a county health department has observed a rise in gastrointestinal illness. The decision is whether to launch a community-wide hygiene campaign or a targeted water quality intervention. The frame must be explicit: "We need to choose between two interventions within the next four weeks, before the peak season." This time-bound, audience-specific frame forces clarity. It also prevents the common pitfall of framing the problem too broadly (e.g., "reduce disease burden") without specifying who acts and by when.

Another pitfall is framing the problem as a solution. For example, starting with "We need to implement a handwashing campaign" skips the diagnostic step. Instead, frame the problem first: "Rates of diarrheal disease in children under five have increased 20% in the past quarter." Then the solution becomes a response to a measurable problem, not a predetermined answer. This distinction is critical for stakeholder buy-in and for avoiding confirmation bias in data collection.

We recommend a simple exercise: write down the decision maker, the decision they face, and the deadline. If you cannot fill all three blanks, the frame is too vague. For instance, "The county health officer must decide by June 1 whether to allocate funds for a hygiene campaign or a water testing program." This frame sets the stage for comparing options and evaluating trade-offs. It also helps you avoid the mistake of assuming a single solution fits all contexts.

Why the Decision Frame Matters

Without a decision frame, problem-solution writing becomes a list of facts without a narrative. The audience does not know what to do with the information. In epidemiology, where data often points to multiple possible actions, the decision frame narrows the focus. It also helps you avoid the trap of "analysis paralysis"—endlessly describing the problem without ever reaching a solution. By committing to a decision and a timeline, you force yourself to prioritize evidence and make a recommendation.

2. The Option Landscape: Three Approaches to Framing

Once the decision frame is set, the next step is to survey the available framing approaches. In epidemiology, three main types dominate: causal framing, descriptive framing, and predictive framing. Each serves a different purpose and carries its own strengths and weaknesses.

Causal Framing

Causal framing asks: "What causes the problem?" It is the classic etiological approach. For example, "Does exposure to contaminated water cause the observed increase in diarrheal disease?" This frame is powerful for identifying interventions that address root causes. However, it requires strong study designs (cohort, case-control, randomized trials) and can be slow and expensive. A common pitfall is assuming causation from correlation, especially when using cross-sectional data. Teams often fall into the trap of overinterpreting observational data without considering confounding variables.

Descriptive Framing

Descriptive framing asks: "What is the pattern of the problem?" It focuses on who, where, when, and how much. For example, "What is the age distribution, geographic clustering, and temporal trend of cases?" This frame is essential for surveillance and outbreak investigations. It is faster and less resource-intensive than causal framing, but it does not directly point to solutions. A pitfall is stopping at description without moving to action. Many reports describe the problem in detail but fail to recommend a solution, leaving decision makers frustrated.

Predictive Framing

Predictive framing asks: "What will happen if we intervene (or not)?" It uses models to forecast outcomes under different scenarios. For example, "If we implement a vaccination campaign, how many cases will we prevent over the next six months?" This frame is valuable for resource allocation and policy planning. However, it depends on the quality of the model and the assumptions built into it. A common pitfall is overconfidence in model outputs, especially when data are sparse or assumptions are hidden. Teams sometimes present predictions as certainties, ignoring confidence intervals and sensitivity analyses.

Choosing among these frames depends on the decision at hand. Causal framing is best when you need to identify an intervention target. Descriptive framing is ideal for situational awareness and communication. Predictive framing is most useful for comparing intervention scenarios. Many projects benefit from combining frames: start with descriptive to characterize the problem, then use causal to identify drivers, and finally predictive to evaluate solutions.

3. Comparison Criteria Readers Should Use

When evaluating which framing approach to use (or how to assess someone else's framing), you need clear criteria. We recommend four criteria: relevance to the decision, feasibility within constraints, risk of misinterpretation, and actionability of the output.

Relevance to the Decision

The frame must directly inform the decision maker's choice. If the decision is about resource allocation, a descriptive frame showing disease burden is relevant. If the decision is about which intervention to implement, a causal or predictive frame is more relevant. A pitfall is using a frame that answers a different question than the one being asked. For example, providing a detailed descriptive analysis when the decision maker wants a causal recommendation leads to misalignment and frustration.

Feasibility Within Constraints

Time, budget, data availability, and expertise all constrain framing choices. A causal study may be ideal but impossible within a two-week outbreak response. In such cases, a descriptive frame combined with expert judgment may be the best option. The pitfall is insisting on an ideal frame that cannot be executed, leading to delays or no analysis at all. Teams should be honest about constraints and adapt their framing accordingly.

Risk of Misinterpretation

Some frames are more prone to misinterpretation than others. Causal frames can be misinterpreted as definitive proof when the evidence is only correlational. Predictive frames can be misinterpreted as forecasts rather than scenarios. Descriptive frames can be misinterpreted as implying causation. The framing should include explicit caveats and uncertainty. A pitfall is presenting results without uncertainty, which can lead to overconfident decisions.

Actionability of the Output

The ultimate test of a frame is whether it leads to a clear action. Does the output tell the decision maker what to do differently? A descriptive frame might show that cases are concentrated in one neighborhood, but it does not say what to do. A causal or predictive frame can recommend a specific intervention. The pitfall is producing elegant analyses that end with "more research is needed" without any actionable recommendation. While further research is sometimes necessary, the frame should still offer provisional guidance based on the best available evidence.

4. Trade-Offs Table: Comparing Framing Approaches

To help you choose, we summarize the trade-offs among the three framing approaches in the table below. Use this as a quick reference when designing your problem-solution structure.

The table reveals that no single frame is universally superior. The best choice depends on the decision context, available resources, and tolerance for uncertainty. A common mistake is to default to causal framing because it seems more rigorous, even when descriptive or predictive framing would be more practical and timely. Another mistake is to use predictive framing without validating the model, leading to misleading forecasts.

5. Implementation Path After the Choice

Once you have selected a framing approach, the next step is implementation. This involves operationalizing the frame into a concrete plan: defining variables, collecting data, conducting analysis, and communicating results. Each framing approach requires a different implementation path.

Implementing Causal Framing

If you choose causal framing, start by specifying the causal question using a structured framework like the PICO (Population, Intervention, Comparison, Outcome) format. For example: "Among children under five in County X, does a water quality intervention compared to a hygiene campaign reduce diarrheal disease incidence over six months?" Then select an appropriate study design. Consider potential confounders and plan for sensitivity analyses. A pitfall is skipping the design phase and jumping to data collection, which can lead to uninterpretable results. We recommend writing a brief analysis plan before collecting data, including a directed acyclic graph (DAG) to visualize causal assumptions.

Implementing Descriptive Framing

For descriptive framing, begin by defining the population, time period, and geographic area. Use standardized case definitions to ensure consistency. Create epidemic curves, maps, and demographic breakdowns. A common pitfall is over-aggregating data, which can mask important patterns. For instance, reporting only overall incidence may hide age-specific spikes. Another pitfall is failing to update the description as new data come in. Descriptive framing should be dynamic, especially during an outbreak. We recommend setting a regular cadence for updating descriptive reports (e.g., daily or weekly) and including both raw numbers and rates.

Implementing Predictive Framing

Predictive framing requires building a model that captures the transmission dynamics or intervention effects. Start by specifying the model structure (e.g., compartmental model, agent-based model) and the key parameters (transmission rate, recovery rate, intervention efficacy). Calibrate the model to available data and validate it against a holdout period. A major pitfall is overfitting—tuning the model to match past data so closely that it fails to generalize. Another pitfall is not communicating uncertainty. Always present predictions with confidence intervals or scenario ranges. We recommend running sensitivity analyses to show how results change under different assumptions.

Regardless of the framing approach, documentation is critical. Record your decisions, assumptions, and data sources. This transparency allows others to assess the validity of your framing and replicate your analysis. It also helps avoid the pitfall of hindsight bias, where you later rationalize choices that were made under uncertainty.

6. Risks If You Choose Wrong or Skip Steps

Poor framing choices carry real consequences. In epidemiology, these can range from wasted resources to harmful public health recommendations. Understanding the risks can motivate teams to invest time upfront in getting the frame right.

Misaligned Resources

If you choose a causal frame when a descriptive frame would suffice, you may spend months and significant budget on a study that only confirms what surveillance data already showed. Conversely, using a descriptive frame when a causal question is needed can lead to interventions that address symptoms rather than causes. For example, distributing hand sanitizer during an outbreak may have little effect if the primary transmission route is contaminated water. The risk is that resources are allocated to ineffective or suboptimal solutions.

Misleading Conclusions

A frame that does not match the data can produce misleading conclusions. For instance, using a predictive model without adequate validation can give false confidence in a forecast, leading to over- or under-preparation. In one composite scenario, a health department used a predictive model that assumed constant transmission rates, but the outbreak was declining due to seasonal factors. The model predicted a surge that never came, causing unnecessary alarm and diversion of resources from other priorities. The pitfall was not updating the model with real-time data and not accounting for seasonality.

Loss of Credibility

When framing is inconsistent or poorly justified, stakeholders lose trust. If a report frames the problem as a crisis but offers a solution that seems mismatched, decision makers may dismiss the entire analysis. For example, framing a small cluster of cases as an impending epidemic (descriptive framing) and then recommending a citywide lockdown (causal solution) can seem alarmist if the evidence does not support the scale of the response. Credibility is built on alignment between the problem frame and the solution frame.

Ethical Risks

Framing can also have ethical implications. A frame that emphasizes individual behavior change over structural factors can unfairly blame affected communities. For instance, framing a sexually transmitted infection outbreak as a problem of personal responsibility ignores social determinants like access to healthcare and education. The risk is that the solution becomes punitive rather than supportive. Teams should consider how their framing might affect vulnerable populations and strive for frames that promote equity.

Skipping steps—such as not defining the decision frame, not considering alternative frames, or not validating models—amplifies these risks. The most common mistake is rushing to a solution frame without adequately defining the problem. This leads to solving the wrong problem or proposing a solution that does not fit. We recommend a mandatory "frame check" before any major analysis: ask whether the problem is clearly defined, whether the frame matches the decision, and whether the assumptions are explicit.

7. Mini-FAQ: Common Questions About Problem-Solution Framing

This section addresses frequent questions we encounter when working with epidemiology teams on framing.

What if the decision maker is not clear about their choice?

This is common. In that case, your framing should help them clarify. Start with a descriptive frame to characterize the situation, then present a few plausible decision options. For example, "Based on the current data, you could either expand surveillance, launch a vaccination campaign, or do both. Here is what each option would likely achieve." This approach moves the decision maker from vague concern to specific choices. Avoid forcing a decision frame that does not exist; instead, use the framing process to co-create clarity.

How do I handle multiple audiences with different needs?

One frame rarely fits all. For a technical audience, you might use a causal frame with detailed methods. For policymakers, a predictive frame with clear scenarios and cost estimates may be better. For the public, a descriptive frame with simple visuals often works best. The pitfall is trying to serve all audiences with one frame, which can confuse everyone. We recommend creating layered communications: a core analysis with a primary frame, then tailored summaries for each audience. The key is to be transparent about the frame you used and why.

Can I change the frame mid-project?

Yes, and sometimes you should. As new data emerge, the original frame may no longer be appropriate. For example, a descriptive frame might reveal a clear causal pathway, prompting a shift to causal framing. Or a predictive model might show that the original intervention is unlikely to work, leading to a reevaluation of the solution. The risk is sticking with a frame that is no longer useful because of sunk cost. We recommend periodic frame reviews, especially after major data updates. Document any changes and the rationale.

What is the biggest mistake teams make?

In our experience, the biggest mistake is framing the solution before the problem. Teams often start with a preferred intervention (e.g., "We should implement a contact tracing app") and then look for a problem to justify it. This leads to confirmation bias and weak evidence. The correct order is: define the problem, then evaluate solutions. Another common mistake is using a single frame when multiple frames would provide a richer picture. Combining descriptive and causal frames, for instance, can be powerful.

How do I know if my framing is working?

Test it with a member of the target audience. Ask them: "What decision does this analysis help you make?" If they can articulate a clear choice, the framing is likely effective. If they say "I understand the problem but do not know what to do," the framing may be too descriptive. If they say "I know what to do but do not understand why," the framing may lack causal depth. Iterate based on feedback. Also, check for alignment between the problem statement and the solution recommendation—they should feel like a natural pair.

8. Recommendation Recap Without Hype

Framing is not a one-time step but a strategic practice that shapes the entire epidemiological investigation. To avoid common pitfalls, we recommend the following specific next moves:

• Start with the decision frame. Before any analysis, write down who decides, what they decide, and by when. This prevents vague problem statements and ensures your work is actionable.
• Choose a framing approach deliberately. Use the criteria of relevance, feasibility, misinterpretation risk, and actionability to select among causal, descriptive, and predictive frames. Do not default to the most familiar or prestigious frame.
• Combine frames when appropriate. A descriptive frame can set the stage, a causal frame can identify drivers, and a predictive frame can evaluate solutions. Layering frames often yields the most robust guidance.
• Document assumptions and uncertainty. Every frame rests on assumptions. Make them explicit and communicate uncertainty honestly. This builds trust and allows others to assess the strength of your conclusions.
• Test your framing with a real decision maker. A quick sanity check can reveal gaps or misalignments. Revise before investing heavily in data collection or analysis.

By following these steps, you will produce problem-solution framing that is clear, credible, and useful. The goal is not to find the perfect frame but to find a frame that serves the decision at hand. In epidemiology, where stakes are high and time is often short, good framing is a skill worth mastering. Apply it consistently, and you will see fewer wasted efforts and more impactful public health outcomes.