Curiosity · Information for Action

Learning objectives

By the time you have read this chapter, you will be able to:

Apply the "Five Ws" (Who, What, When, Where, Why) to any routine health data set, transforming raw numbers into structured questions that lead to local action.
Calculate and interpret proportions, percentages, rates, and ratios to make fair comparisons between facilities, time periods, or population groups.
Use epidemiological thinking to investigate a suspected outbreak by verifying data, characterising by time/place/person, formulating a hypothesis, and taking local action.
Build a culture of curiosity in your team by embedding data questions into routine meetings, encouraging "but why" inquiries, and connecting data to field visits.

Introduction

Data sits in your RHIS like a patient waiting for a diagnosis. It has a story to tell, but it will not speak unless someone asks the right questions. That someone is you. The tools you need are not complex statistics or expensive software: they are curiosity and a structured way of thinking.

This is the essence of epidemiological thinking. It is the disciplined, systematic practice of asking who, what, when, where, and most importantly, why about the health of your population. It transforms you from a passive recipient of reports into an active investigator of your community's health.

For managers at the facility, district, and beyond, epidemiological thinking is the engine that powers the Information Cycle introduced in Chapter 2. It provides the questions that drive data collection (Stage 2), the analytical framework that turns data into information (Stage 3), and the intellectual curiosity that fuels the discussion and interpretation where true insight is born (Stage 5).

This chapter is an invitation to cultivate your curiosity. We will explore how to use the RHIS not just to count, but to understand; to move from seeing numbers to seeing patterns, risks, and opportunities for action. By the end, you will see your data not as a burden, but as a map, one that reveals where your community is healthy, where it is suffering, and where you, as a manager, can make the greatest difference.

1. The cornerstone: Why epidemiological thinking matters

Epidemiology is often defined as the study of the distribution and determinants of health-related states in populations. But for a district health manager, it is simpler than that: it is the science of asking good questions about your community's health.

This mindset matters because:

It focuses on populations, not just individuals. A clinician treats one patient at a time. A manager using epidemiological thinking asks: What is happening across all 50,000 people in my district? Who is being left behind?

It quantifies the problem. You cannot manage what you cannot measure. Epidemiological thinking gives you the tools (rates, ratios, proportions) to move from vague impressions ("malaria seems bad this year") to precise statements ("malaria incidence in under-fives has increased by 40% in the northern clinics compared to last year").

It looks beyond the health system. Most illness is driven by factors outside the clinic: water quality, poverty, education, behaviour. Epidemiological thinking pushes you to connect your RHIS data to these underlying social and environmental determinants.

In short, epidemiological thinking is the bridge from routine clinical data and derived indicators to strategic public health intelligence. It is what allows a monthly statistical report to become a compelling story that can change policy, move resources, and save lives.

2. The five ws: The questions that unlock your data

Every good investigation starts with a set of fundamental questions. For the epidemiological thinker, these are the "Five Ws." Your RHIS is designed to capture data that can answer them. The challenge is to make answering them a routine part of your management practice.

Figure 4.1: The Five Ws: the questions that unlock your data.

Epidemiological questions
WHO?	Which populations are affected? Age, gender, occupation, risk groups. Numbers of cases by age group or sex. Unique patient identifiers allow for tracking of individuals, co-morbidities, and care pathways.
WHAT?	What are the main conditions or health events? Aggregated counts of diagnoses (e.g., 50 cases of pneumonia). Specific disease names, codes, and details of interventions provided.
WHEN?	What are the temporal patterns? Seasonality, trends over months or years. Data is normally aggregated and reported by month. Real-time, daily, or even hourly data allows for rapid outbreak detection.
WHERE?	Where are the problems concentrated? Geographic hotspots, specific facilities, villages. Data is linked to the reporting facility. Georeferenced household-level data allows for precise mapping of cases and risk factors.
WHY?	What are the underlying causes? This is the most complex question. Routine data alone rarely answers "why." It raises the question. Linking RHIS data to external databases (e.g., water sources, weather patterns) can reveal associations.

The first four Ws (Who, What, When, Where) are the foundation. They tell you what is happening. A manager who masters these can identify a spike in diarrhoea (What) in children under five (Who) during the rainy season (When) in the northern part of the district (Where).

But the final W, Why, is where the true power of epidemiological thinking lies. It is the leap from description to understanding. It requires you to look beyond the data in your RHIS to consider:

Social determinants: Is poverty driving this pattern?

Environmental factors: Is there a contaminated water source in that northern area?

Behavioural patterns: Are there cultural practices or beliefs influencing care-seeking?

The RHIS may not directly answer "why," but it is the starting point. It tells you where to look and what questions to ask when you go out into the community, which is a central part of answering the WHY. Often it is the story of what is happening in the community that discloses the WHY....

3. Epidemiological thinking: From curiosity to action

Cultivating curiosity is not an abstract exercise. It has concrete, practical applications that will make you a more effective manager.

3.1 Detecting and investigating outbreaks

Example 1

A sudden increase in a single indicator should trigger your curiosity. Why are malaria in children cases spiking in Clinic X when they are stable everywhere else?

This is where your epidemiological toolkit comes into play:

Verify the data: Is it a real increase or a data quality issue? (Trust but verify)

Characterise by time, place, and person: Create a line list. Map the cases. Who is affected? When did they fall ill? Where do they come from?

Formulate a hypothesis: Is it a new mosquito breeding site following heavy rains? A delay in insecticide-treated net distribution? An influx of non-immune populations?

Investigate and act: Deploy a team to the field. Test the hypothesis. Implement control measures, drain the breeding site, distribute nets, spray the houses, activate community health workers for active case finding.

Your curiosity, sparked by a routine RHIS report, has now led to a direct public health intervention.

Investigate and act: Form a group discussion with these young girls. Discover that there is a new construction site being developed and there have been more men with disposable income. Action includes increasing condom distribution, providing emergency contraception and drugs to prevent HIV transmission. Create an adolescent group that meets weekly also including family planning service and STI treatment in a more contained area to avoid everyone knowing private details about the adolescents.
Formulate a hypothesis: There has been a change in socio-economic conditions, with more young adults having an earlier sexual debut.
Characterise by time, place, and person: Map the cases. When did this start? Where are these pregnant women coming from, is it spread around, or limited to one area
Verify the data: Is it a real increase or a data quality issue? (Trust but verify)

There is a spike in Antenatal clients HIV positive rate in the age group 15 to 24 over the last 6 months.

Example 2

EXERCISE 3: Outbreak investigation (the Fig malaria spike)

Part A: Fruit District Scenario

You have verified the Fig malaria data (it is real; the lab confirmed the RDT results). You have sent Mr. Guava on the motorcycle. He calls you from Fig:

"Dr. Mango, I am at Fig. The cases are all from two villages, Mkuu and Mto, both near the river. Most are children under 10. The nurse says the CHW for Mkuu resigned two months ago and was never replaced. There is a new stagnant pool near the Mto bridge; the road construction crew dug it and left it. Also, the last net distribution missed Mkuu because the CHW was gone."

Task: Walk through the four-step outbreak investigation process:

Then answer:

What local action can you take without waiting for national approval or additional budget? (The rains start in two weeks.)

What will you do about the stagnant pool? The missing CHW? The nets?

What will you do differently for Date and Cherry (the other two facilities that become inaccessible) based on what you learn from Fig?

Part B: Apply to your own facility or district

Now think about a recurring problem in your own facility or district that might benefit from an outbreak investigation mindset. It does not have to be an infectious disease. It could be a sudden drop in immunisation coverage, a spike in patients missing appointments, a repeated stockout of a specific drug, a rise in patient complaints about waiting time....

Apply the same four steps to your chosen problem and create the same table of epidemiological actions.

PDSA: after you act, study. Keep watching the weekly lab data. Did cases drop in other facilities? If no, ask what you missed. Something else? Use what you learn to act again. An outbreak investigation is not a one-off event. It is a cycle of action, observation, and adjustment until the problem is contained.

Output for Part B: A one-page "Outbreak Investigation" (even if it is not an outbreak) documenting your four steps and your committed local action.

3.2 Pursuing health equity

Epidemiological thinking forces you to look beyond averages. A district-wide immunisation coverage of 85% might hide a reality where one facility has 95% coverage and another has only 60%.

By asking who is not being served, where they live, and why they are not accessing services, you can move from a general problem to a targeted solution. Perhaps the low-coverage area is a remote community without a functional outreach programme. Your job is then not to scold the facility manager, but to work with them to design a mobile outreach schedule or strengthen the network of community health workers.

Curiosity about disparities becomes the engine for equity.

3.3 Connecting health services to health problems

A key part of your role is to ensure that the health services you manage are aligned with the health problems your population faces. This requires another set of "Five Ws" focused on your own health system (see D Werner picture):

By mapping your services against your disease burden data, you can identify gaps and inefficiencies. For example, if the highest rates of teenage pregnancy are in a specific ward with no youth-friendly services, your data has just made a powerful case for a new intervention.

4. The language of epidemiology: Making comparisons

To ask good questions, think like a manager and speak the language of comparison.

Raw numbers, like "50 cases of diabetes", are meaningless without context. Is 50 a lot? For a population of 500, yes. For a population of 5 million, no.

Epidemiologists use three simple tools to make fair comparisons between different months, facilities, or districts: proportions, rates, and ratios.

Term	Description	Example
Proportion	Numerator is part of the denominator. Tells what share of the problem pneumonia represents. Usually expressed as a percentage.	Proportion of under-five deaths due to pneumonia. (Child with pneumonia death (0 to 4 years)/ Death (0 to 4 years) x 100.
Rate	Numerator is part of the denominator. Time-related frequency of an event in a population, this can measure risk	Malaria incidence rate. (New malaria cases in a month / Population at risk) x 1,000.
Ratio	Compares two unrelated quantities. The numerator is not part of the denominator.	Nurse workload ratio. (Patient visits in a month / Nurse-days worked) Bed occupancy. Beds occupied/ Beds available.

Mastering these three concepts allows you to move from asking "how many?" to asking "how many, relative to something else?" This is the heart of comparative analysis and the foundation for setting priorities.

Rates to real numbers for facilities

At the same time we need to convert percentages into numbers so that facility staff understand the actual numbers of clients needed to achieve improvement.

For example, in a facility serving 10,000 people, there are ~400 children <1 yr. If the current coverage is 55% (210 children a year or 18 a month) and you want to increase to 75%, this means that you have to do 300 immunisations, an additional 90 children a year or 8 additional immunisations a month.

EXERCISE 5: Proportions, rates, and ratios

Part A: Fruit District Scenario for 2025

Task: Calculate for each facility:

Percentage IPT 1st dose coverage (numerator ÷ denominator) X 100

IPT 1st dose coverage rate (numerator ÷ denominator × appropriate time period)

Ratio ANC / nurse day worked (numerator ÷ denominator (different than numerator)

Then answer:

Which facility has the highest raw number of IPT doses?

Which facility has the highest proportion (coverage)?

Which facility has the highest productivity (ANC per nurse-day)?

If you had one extra nurse to allocate, which facility would get them? Why?

If you had a small budget for community outreach, which facility would get it? Why?

Part B: Apply to your own facility or district

Now select one action indicator from your own facility (e.g., immunisation coverage, ANC 1st visit, TB cure rate). Collect or estimate: Numerator, Denominator and second denominator for a ratio (e.g., number of staff, number of hours, number of outreach visits)

Then calculate:

Proportion (numerator ÷ denominator × 100)

Rate (numerator ÷ denominator × appropriate multiplier)

Ratio (numerator ÷ second denominator)

Then answer:

What does the proportion tell you about coverage?

What does the rate tell you about risk or service use?

What does the ratio tell you about efficiency or workload?

Which of these measures is more useful for daily management? Which is more useful for annual planning?

Output for Part B: A short memo to your team titled "Our [Indicator Name]: What the Numbers Really Mean", showing your calculations and one conclusion.

4.1 A distinction that matters: Incidence vs. prevalence

Two specific rates are fundamental for managing a district health system:

Incidence: New cases of a disease in a population over a specific period. It tells you about the risk of becoming ill. A high incidence rate suggests a need for prevention programmes. (e.g., a spike in new HIV diagnoses might indicate a gap in prevention efforts). It is the best way to track quick moving diseases such as flu, COVID-19 and measles and we can track the impact of interventions that bring rapid change such as vaccinations.

Prevalence: Existing cases (both new and old) at a specific point in time. It tells you about the burden of existing disease, especially chronic diseases. A high prevalence rate suggests a need for treatment and care services and supports programme planning. (e.g., a high prevalence of hypertension in the community means you need strong chronic disease management programmes).

Think of it like a bathtub:

Incidence is the water flowing in from the tap (new cases).

Prevalence is the amount of water in the tub (all existing cases).

Death or cure is the water flowing out the drain.

Figure 4.3: The bathtub analogy: incidence flows in, prevalence is the level, death or cure drains out.

Understanding this distinction helps you allocate resources appropriately. An effective intervention might lower incidence (preventing new cases) while prevalence remains high for years (as people live longer with a chronic condition), or it might lower prevalence by curing existing cases.

5. A culture of curiosity: Making epidemiological thinking routine

Epidemiological thinking is not an event; it is a habit. It is the culture of curiosity and critical thinking in a health system that is constantly learning and improving.

How do you, as a manager, cultivate this culture?

Start Every Meeting with a Question: Instead of just reviewing reports, ask the team: What surprised you in this month's data? What pattern do you see? What question does this raise for us?
Use the Five Ws as a Checklist: When reviewing a dashboard, consciously go through the questions. Who is affected? What is the condition? Where are the hotspots? When did this start? Why do you think this is happening? Make it a mental routine.
Encourage "But Why" Questions: When a staff member presents a finding, ask them why they think it is happening. Even if they are not sure, the act of hypothesising is the first step to understanding.
Go Beyond the Facility Walls: Use your data to prompt field visits. If the data shows a problem in a specific area, go there. Walk the community. Talk to community health workers and residents. Use your eyes and ears to find the story behind the numbers.
Celebrate the Investigators: Recognise and reward innovative staff who use data to identify problems and propose solutions. Make curiosity a valued trait.

By embedding these practices, you shift your team from being data reporters to being public health detectives. You build a team that does not just manage the health system: they actively investigate it, understand it, and continuously improve it.

EXERCISE 7: Build a culture of curiosity: The Lemon staff meeting

Part A: Fruit District Scenario

Lemon Health Centre's monthly data review consists of the manager reading numbers aloud while staff look at the floor. No one asks "why." Staff are already overworked and see data meetings as "extra work." No one wants to be seen as criticising the manager.

Dr Mango the district manager genuinely wants to change but does not know how.

Task: Put yourself in Dr Mango’s shoes. Design a 30-minute weekly data huddle for Lemon.

What day and time will it happen? (Piggyback on an existing gathering, e.g. the Monday morning handover.)

Who attends? (Mandatory? Voluntary? Rotating? Perhaps the manager, the nurse in charge, and one rotating clinical staff member?)

What is the first question the manager asks to open the meeting? (It should NOT be "What are the numbers?" It should be something that invites curiosity, like "What in last week's data surprised you?" or "Which number made you worry?")

What is the one rule for the meeting? (Example: "No blaming. Only problem-solving." Or "Every question must start with 'Why do you think... ?'")

How does the manager close the loop and show staff that their questions led to action? (e.g., "Last week you asked about X. Here is what we did: Y. Here is what happened: Z.")

Then answer:

- What is one specific question you will ask at your next team meeting to invite curiosity?

- What will you do with the answers to that question?

Part B: Apply to your own facility or district

Now answer these questions for your own team:

Does your team currently have a regular meeting where data is discussed?

If yes: What percentage of the meeting time is spent with one person talking and others listening? What percentage is spent asking "why"?

If no: What existing meeting could you add a 15-minute data discussion to?

What is one question you will ask at your next team meeting to invite curiosity?

What will you do with the answers to that question?

Output for Part B: A one-paragraph "Curiosity Commitment" stating what you will do differently at your next team meeting. Then do it. Then write down what happened.

Conclusion

Your RHIS is a powerful tool, but it is just a tool. The true engine of a data-driven health system is the curious, epidemiological, critical-thinking mind that uses double-loop learning. It is the managerial mind that looks at a table of numbers and sees a story; that looks at a map of cases and sees a family or community at risk; that looks at a trend line and asks, "What can we do differently?"

This chapter has given you the framework: the Five Ws, the language of rates and proportions, the power of incidence and prevalence. But the most important ingredient is one you already possess: curiosity and critical thinking. Nurture it. Apply it to your data. Let it guide your self-assessment, your supervision visits, your planning meetings, and your resource allocation.

When you combine the power of the RHIS with the discipline of epidemiological thinking, you stop being a passive collector of information and become an active, curious, and effective leader of your community's health. You unlock the true potential of your data: not just to count, but to understand; not just to report, but to act; not just to manage, but to heal.

CAPSTONE EXERCISE: The curiosity investigation

Task: You have been asked to present a 5-minute "data story" at your next staff meeting or district review. Choose one indicator from your facility or district that tells an interesting story.

Using everything from this chapter, prepare a one-page story outline that includes:

The hook: What caught your attention? (One sentence.)
The Five Ws: Who?, what?, when?, where?, with specific numbers.
The comparison: A rate, proportion, or ratio that makes the numbers meaningful.
The "why": Your best hypothesis (even if not proven) about what is driving the pattern.
The action: One specific local action you are recommending based on this story.

Then: Treat that action as a small test. Plan how you will know if it worked within 30 days. Tell your audience: "We will try this for one month, study the results and then report back on what we learned."

Output for Capstone: One page. Practice telling it out loud in under 5 minutes. Present it to a colleague and ask: "Did this make you curious?"

References

Vaughan, 1989
Vaughan, 1989
Unger, 1992
Nakuru County study, 2025
Universitas Airlangga (UNAIR), 2025
Vaughan, 1989
Vaughan, 1989
Unger, 1992
World Health Organization, 2023
MEASURE Evaluation, n.d.