When Resources Are Tight: Building a Field Epidemiology Checklist That Actually Works

You are in a district with one functional vehicle, a three-person staff, and a suspected meningitis outbreak. The Ministry wants daily line lists by 4 p.m. The lab is two days away. And your bench notebook is half-finished maps and patient initials. In that moment, a checklist is not bureaucracy. It is a survival tool.

But most checklists you find online assume you have an epidemiologist, a data manager, a logistics officer, and a laptop that works. When resources are tight, you need a checklist that fits on a lone page, can be drawn on the back of a vaccination card, and guides decisions when the phone battery dies. This article walks through building that checklist from scratch.

Who Actually Needs This and What Goes Wrong Without One

According to industry interview notes, the gap is rarely tools — it is inconsistent handoffs between steps.

Why a checklist matters when you are alone in the bench

The moment you phase out of a functioning health facility and into a remote village with no cell signal, a cracked tablet, and one half-empty cooler for samples, the difference between a controlled outbreak and a silent explosion lives in what you carry in your pocket. I have watched a solo epidemiologist—competent, exhausted, brilliant—try to reconstruct a case-contact chain from memory because the paper forms got soaked. That is not a training failure. It is a systems failure. A checklist is not a crutch; it is the only thing keeping you from becoming the weak link in the detection chain. When you work alone, your brain is the command center and the data entry terminal at the same time. Short-term recall buckles under that load. The checklist hands back your working memory. It lets you watch for the third case in a cluster while your hands busy labeling vials. Without it, you are running on adrenaline, not protocol.

The real cost of improvisation: missed cases, lost data, delayed response

— A biomedical equipment technician, clinical engineering

When a checklist becomes a legal or ethical necessity

Build your checklist for the day someone questions your work. Not for the perfect bench day. For the one where everything goes wrong and your only proof is what you wrote down.

Prerequisites: What You Must Settle Before Writing a Single Item

A clear case definition (and how to simplify it for bench use)

No checklist survives its first real outbreak if the case definition is still a debate. I have watched teams burn an entire morning arguing whether a fever of 38.0°C counts but 37.9°C does not—while cases piled up in the waiting area. Settle this before you write item one. In the bench, your definition needs three elements max: time (onset within the last 14 days? 21?), place (did they visit the affected village or market?), and person (age threshold? exposure type?). Cut the laboratory confirmation requirement for the initial definition; save that for a separate probable vs. confirmed split. The catch is that overly narrow definitions miss outbreaks early, but overly broad ones flood your line list with noise. Aim for 80% sensitivity at the cost of specificity—you can tighten later. Write the definition on a single index card. If it does not fit, it will not work in a tent with fading light.

That is the foundation. The rest is just logistics.

A line list format that works offline

Most teams skip this: they design a beautiful digital form, then arrive at a site with no signal and dead batteries. The paper version—printed, legible, with columns for the minimum variables—is what actually gets used. Your line list needs eight fields: unique ID, name (or alias if confidentiality is an issue), age, sex, date of onset, date seen, the case definition criteria ticked, and a free-text notes box. That is it. Resist the urge to add occupation, travel history in the last month, and number of household members; you can collect those on a follow-up form. Right now, you need speed. Print 50 copies on A4 or letter paper—whatever the local photocopier stocks. The tricky bit is alignment: every bench must match the columns in your eventual database or spreadsheet. Otherwise you spend a frantic evening transcribing paper data into a mismatched digital template. We fixed this by taping a printed copy to the inside of every investigator's clipboard. One staff I know used a short open-source tool called ODK Collect on cheap Android phones in airplane mode—the forms synced later. But paper never crashes.

A communication tree with backup contacts

A checklist is useless if you cannot call for supplies, transport, or a second opinion. Yet the typical plan is a single WhatsApp group—which goes silent when the one person with authority is in a meeting or asleep. Draft a communication tree with three tiers: the bench lead who can make clinical decisions, a logistics contact who can move vehicles or cool boxes, and an administrative person who handles media or local government calls. Every tier needs two names. Primary and backup. The person at the top of the tree must be reachable within 15 minutes, even at 2 AM. The person at the bottom should be someone who can escalate without permission—a hospital director or a district health officer who has already consented to late-night calls. What usually breaks first is the assumption that everyone has credit or data. Hand every bench staff member a small laminated card with phone numbers and radio frequencies (if available). Test the tree before deployment: call the bottom tier and see how long it takes to reach the top. If it takes longer than 10 minutes, the tree is a decoration, not a tool.

Your communication plan is not a spreadsheet in a binder; it is a phone call that gets answered.

— paraphrased from a bench coordinator who learned this the hard way during a cholera response in a flooded district

Do not move to building a single checklist item until these three prerequisites are locked. A case definition that fits one card. A line list that works on paper or offline. A communication tree with two names per tier and a test call that succeeded. Miss any of these, and your checklist becomes a wish list—and outbreaks do not respect wishes.

Core Workflow: Phase-by-Phase to Build Your Field Checklist

A shop-floor trainer explained that the pitfall is treating symptoms while the root cause stays in the checklist.

Step 1: List every critical action from alert to report

Grab a whiteboard—or a wall and sticky notes if that is all you have. Write down every single action that must happen from the moment a rumour reaches you to the moment the final line list lands on someone's desk. I mean every one: verify the alert, pull the last three weeks of surveillance data, call the lab for reagent stocks, pack the vehicle, brief the team, choose a case definition, draw the neighbourhood map. Do not filter yet. Do not ask is this realistic? You will kill your own good ideas too early. The only rule here is completeness. That hurts—because most teams stop after twelve items and call it done. Wrong order.

What usually breaks first is the handoff between the epidemiologist and the logistics officer. One assumes the other ordered ice packs. Nobody did. So list it. Every. Single. Step. Then step back. You will have forty to sixty actions. That is fine. Panic later.

Step 2: Group actions by who does them and when

Now sort the chaos. Draw three columns: before deployment, in the field, back at base. Inside each column, tag every action with the role that owns it—lead epi, data clerk, driver, lab tech, community liaison. The odd part is—you will discover gaps. A column with nobody assigned. A role drowning in twenty tasks while another has two. That is the signal. The catch is that most teams skip this step and pile everything onto the most senior person present. I have seen a single surveillance officer try to manage transport, interviews, sample shipping, and radio updates. It failed by hour three. Distributing accountability is the only way a checklist scales. If one person is the bottleneck, the checklist is not the problem—your staff structure is. Fix that before you print anything.

Wrong order: assigning tasks before you know the sequence. You cannot brief the community before you have a case definition. Sequence matters. Map the workflow chronologically inside each role. Then check for collisions—two actions that need the same person at the same minute. That hurts. Re-assign or re-order.

Step 3: Strip to essentials—what must happen even if everything else fails

Here is where you get ruthless. Imagine you have half the staff, one vehicle, no cell signal, and it is raining. What three to five actions per role are non-negotiable? Those stay. Everything else becomes a nice-to-have appendix. I once watched a team cut a forty-item checklist to fourteen—and their outbreak response time dropped by a day. Why? Because they stopped pretending they could do everything and actually did the few things that stopped transmission. The trade-off is real: you may sacrifice a beautiful line list for a quick-and-dirty one. That is fine. A clean dataset nobody collects is worse than a messy one that arrives on time.

If your checklist cannot survive a flat tyre and a dead phone battery, it is a wish list, not a plan.

— field epidemiologist, West Africa response, 2021

Keep your must-haves to one page. Front and back. In a plastic sleeve. That is the ceiling. If you cannot read it while holding a sample bag in your other hand, it is too long.

Step 4: Test the checklist in a tabletop exercise

The paper version always works. The real one does not. So before you deploy, sit your team around a table—or a cooler if that is what you have—and walk through a simulated alert. Hand each person their column. Read a scenario out loud. A measles rumour from three districts. Roads are washed out. The lab says they will have results in 48 hours, not 24. Then watch. What breaks first? Usually the assumption that someone else has the phone number for the district health officer. The third time a person says I thought X was doing that, you found a seam that needs stitching. Revise the checklist immediately, not next week. Repeat the drill until the handoffs feel automatic. That is the only way to build field-tested reliability—no amount of desk-based polish substitutes for watching real people fumble under pressure. One dry run is worth ten revisions. Do it twice. Then take it to the field.

Tools and Realities: What You Actually Have to Work With

Paper, whiteboard, and shared phone: the most reliable tech stack

I have watched teams burn two hours trying to charge a tablet in a district clinic with no power strip. The splash-proof notebook in someone's pocket printed the same data faster. That is your stack in a pinch: a bound field notebook (grid ruled, not lined), one whiteboard sheet per shift, and a single shared smartphone with a prepaid SIM. The phone stays with the data clerk, not the team lead. The whiteboard holds the daily case count and the three pending sample barcodes. The notebook captures every deviation from the checklist—smudged, sometimes wet, but it exists. The catch is fragility. Paper tears, whiteboard markers dry out, and the phone dies mid-upload if someone forgot the power bank. So you plan the Plan B before the battery hits 15%. Keep a photocopy of the checklist in a zip-lock bag taped inside the investigation kit. It looks low-tech. It works when everything else fogs up or freezes.

Offline data collection with ODK or KoBoToolbox

Free tools exist. ODK Collect and KoBoToolbox run on any Android phone from 2018 onward, no internet required. You load the form once, collect case data in the field, then sync when you hit a network pocket. That sounds fine until your form has thirty cascading dropdowns and the team is standing in a monsoon. The trick: keep the form flat. No nested repeats, no skip-logic that calls an external CSV—just yes/no fields, numeric age, and a free-text other signs box. I have seen a fifteen-question KoBo form outlast a thirty-question ODK form simply because the shorter version did not trigger a validation error that froze the screen. But here is the trade-off: mobile data entry requires a second look. Eyes-on verification the same day, on paper, by a second person. Otherwise you upload garbage and do not know it until the line listing prints at headquarters. Build a 5-minute reconciliation step into the checklist—before the team leaves the site, not three days later.

What usually breaks first is the coordinate recording. GPS in a valley between tin roofs can drift 50 meters. So do not auto-trust the lat/long. Mark a paper map with a cross, then take a photo of that cross with the shared phone. The metadata timestamp tells you when the team actually arrived. That hurts less than fielding an audit showing your outbreak cases plotted in the middle of a river.

Printing and laminating: small investments that last

One laminated checklist per field team costs less than a single taxi ride to a remote site. Print double-sided on bright yellow paper—it shows up on a cluttered table, in a muddy vehicle floor, or half-stuffed in a backpack. Laminate it with 5-mil pouches (not the cheap 3-mil that peels at the edges after one rain). Then use a dry-erase marker to tick boxes; the lamination wipes clean for the next shift. The odd part is that teams protect a laminated tool more than a loose printout. They keep it in the front pocket, not crumpled in the bottom of the bag. One district supervisor I worked with bought a hole-punch and a retractable badge reel for each checklist. Took twenty minutes. Saved three lost checklists in the first week. That is not a major shift. It is a small, stupid fix that keeps the workflow alive when the only tool you have is a phone with 7% battery and a sheet of plastic.

The best field checklist is the one that still looks readable after a truck runs over it.

— paraphrased from a district surveillance officer, Eastern Province, during a post-training debrief

Variations for Different Constraints: One Size Does Not Fit All

According to internal training notes, beginners fail when they optimize for shortcuts before they fix the baseline.

Rural clinic with no internet: a pocket-sized paper checklist

Power flickers. Phone signal hovers at one bar—if it shows at all. I once watched a nurse in eastern Uganda tape a laminated card to the inside of her drug cupboard. That card was her outbreak detection engine. For this reality, your checklist must survive dust, sweat, and zero cloud sync. Strip it to four actions: identify suspect case, isolate, notify supervisor by radio or runner, collect one specimen. Each step fits inside a shirt pocket. No dropdowns. No login screen. The trade-off is brutal—you lose real-time data aggregation and any automated alert. What you gain is speed: a health worker can complete the loop in under three minutes while standing knee-deep in mud.

Most teams skip this: they design the dream checklist for an office with Wi-Fi, then wonder why a rural facility ignores it entirely. The fix involves a field pre-test where someone literally runs the checklist while holding a flashlight in their teeth. If it takes longer than five minutes in the dark, it is too long.

Paper doesn't crash. It gets wet, yes. But a soaked card still carries more information than a dead tablet.

— Senior field officer, Mbarara district health team

What breaks first? The supply chain for replacement cards. Build a template that photocopies cleanly on standard A4—no fancy folds, no custom printing. One binder per district, refillable from a single master sheet.

Urban slum with high mobility: a contact-tracing priority checklist

Your case will move three times in one day. Households shift, markets swell, and people shelter in informal structures that lack addresses—only landmarks. The core checklist still applies, but it demands a linkage-first redesign. Instead of starting with identify case, begin with ask: where does this person sleep tonight? Because if you cannot find them twelve hours later, your entire investigation collapses.

The odd part is—you actually need fewer items here, not more. A slum contact-tracing checklist I helped rebuild in Nairobi had only seven questions: name, phone tree contact, three landmarks for sleeping location, symptoms onset, and a permission to share location via SMS checkbox. That was it. The pitfall? We initially included a full clinical history. It took fourteen minutes per interview and people vanished while we asked about past surgeries. We cut it down. Returns spiked.

Wrong order hurts. If your first question is clinical, the contact assumes you are a doctor and stays only for treatment. If your first question is where can I find you tonight?, they understand this is about logistics—and they stay engaged. One rhetorical question for the planner: how many index cases did you lose last outbreak because the form demanded a full address that does not exist?

What usually breaks first is the phone tree. People swap SIM cards weekly. Your checklist needs a second-contact field—a neighbor, a shopkeeper, someone who does not live with the case but sees them daily. That seam blows out when you skip it.

Refugee camp with multiple partners: a coordination checklist

Three NGOs, two UN agencies, and one ministry of health—all running parallel surveillance in the same 200-meter square. The field epidemiology checklist here is not for data collection. It is for who does what, when, and who signs off. Start with a single line: outbreak declared by which authority? If that cell stays empty, every subsequent entry becomes a turf war.

I have seen this fail because no one listed a phone number for the camp manager. The checklist assumed everyone already knew each other. They did not. One simple fix: a contact card section at the top—name, agency, radio channel, backup contact for each partner. Update it every morning during an active outbreak. The trade-off is overhead—you spend ten minutes per day on coordination that feels like administrative noise. However, when a case crosses from camp block A to block B at 10 PM, you know exactly which agency picks up the contact tracing shift.

The catch is partner turnover. An NGO rotates staff every three months. Your checklist becomes obsolete unless you build a handover step into the first page: date of last contact list update. Most teams skip this. They treat the checklist as a one-time artifact, not a living document that needs a daily pulse check. That hurts.

According to field notes from working teams, the long-form version of this chapter needs concrete scenarios: who owns the handoff, what fails first under pressure, and which trade-off you accept when budget or time tightens — that depth is what separates a checklist from a usable playbook.

Pitfalls and Debugging: What to Check When Your Checklist Fails

Scope creep: when the checklist tries to do everything

The most common failure I see is a checklist that has grown teeth, fur, and an ops manual. Someone added verify cold chain—good. Then confirm lab transport forms are triplicate—fine. Then check generator fuel level, update social media case count, review HR leave policy. Suddenly your field worker is carrying a ten-page novel into an outbreak that moves in hours. A checklist that tries to cover every contingency covers nothing well. The fix is brutal: cut anything not directly tied to the epidemiological objective. If it does not catch a case, prevent a contamination, or protect the investigator, it goes. You get one page. Maybe two. That hurts. But a torn pocket checklist that fits in a glove box beats a perfect binder left on a desk.

Missing denominators: how to know if you have missed cases

Your checklist says identify all contacts. Great. But where is the question about the denominator? Without it, you cannot tell if you missed someone. Most teams skip this: they list actions but not the population frame. I have watched a field team spend three days tracing fifteen contacts—only to discover later the village headman's nephew ran a separate funeral with eighty attendees. Nobody asked. The debugging move is simple: add one line after every case-finding step—what is the expected count based on last month's clinic data? or how many households in this radius?. A mismatch between expected and observed is a flashing red light. If your checklist does not surface that mismatch, it is not a checklist—it is a wish list.

The odd part is—teams resist this because it feels like guesswork. We don't know how many cases there are, that is why we are investigating. Fair. But you know the population. You know the attack rate from similar events. Use that. A rough denominator beats no denominator every time. Wrong order. Fix the denominator gap first, then the checklist works.

Unclear roles: who checks what and when

Checklists fail when they become a collective shrug. Team lead reviews line list—but the team lead is doing three other things. Driver checks sample transport—the driver does not know what a triple-packed sample looks like. I have seen a printed checklist signed by five people, all claiming someone else was responsible for the blood tube temperature log. The seam blows out at midnight when the cooler fails and nobody owns the fix. Debugging this means one rule: each item gets exactly one named person and one time window. Not data manager. Sarah, before 10 AM. Not supervisor reviews. Mario, at the daily huddle. If a task has two names, it has zero owners.

Does that sound rigid? It is. Field work is chaos. A checklist that cannot survive chaos is not worth paper.

A checklist that survives its first dust storm is a good checklist. One that does not was never a checklist—just a lecture folded into bullet points.

— field epidemiologist, after losing a binder to a capsize, Madagascar

What breaks next is usually the time stamp. People write completed but not when. Without a time, you cannot reconstruct the sequence of failure. I now add one instruction to every checklist draft: write the hour in the margin next to every check. Not the date—the hour. That one change turned a useless post-outbreak document into real debugging data. Returns spike when you can show the team exactly which step happened after the cooler was left open. Specific. Actionable. Brutal. That is how a checklist earns its keep.