Week 2 Topic discussion 2

[Lokachet Tanasugarn]

Omoleke and de Kiev’s mixed-methods evaluation found the AEFI surveillance system in Kebbi State, Nigeria, to be suboptimal. Key findings showed convergence on simplicity and timeliness, partial agreement on several other attributes, dissonance on representativeness, and concerns about completeness, timeliness, knowledge gaps, and stability. The authors cautioned that the system is not robust enough to generate convincing safety data for new vaccines.

1.)  Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?

– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?

[Wah Wah Lwin]

1. Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?

The single design limitation that most threatens valid estimates of sensitivity and representativeness of the AEFI surveillance system is the limited geographic and facility coverage, specifically the under-reporting from private and rural health facilities. This may lead to weak data collection, incomplete data reporting, and a lack of representativeness for the coverage.

To address this within six weeks, we need to focus on integration and collaboration with private and rural health facilities for the AEFI surveillance system. Firstly, we need to organize short compulsory training sessions (1-2 days in week one) for key staff from the above health facilities that offer vaccinations. This training will ensure everyone understands the AEFI case definitions and knows exactly how to fill out the required reporting forms, addressing critical knowledge gaps. Secondly, we need to establish clear guidelines that every confirmed adverse event must be reported immediately. This will create a clear, fast channel for data flow and speed up the response (within the timeframe). Finally, we need to conduct review meetings on AEFI reporting, along with acknowledgment of the health facilities that report AEFI cases timely and consistently with good data quality (in week six). This would help the AEFI frontline staff stay motivated, feel recognized, and gain a sense of ownership.

2. Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.

Since sensitivity is sub-optimal due to high under-reporting, delayed transmission, and low community awareness about AEFI and its reporting, a low-cost intervention would be to increase community awareness by conducting education and outreach activities using communication tools such as pamphlets, flyers, and community talks. This aims to empower the community to initiate passive reporting. At the same time, immunization staff will provide education during their visist for immunization, on how to report AEFI cases through simple method such as sending SMS messages for AEFI cases.

Expected trade-offs: Improved data quality and greater acceptability of the system by the community which will help address under-reporting, delayed transmission, and low public awareness.

Indicators to detect the impact of the intervention (based on simplicity and generalizability):
• Increased reported AEFI cases: This will be calculated as the ratio of AEFI reports per 100,000 surviving infants per year. Reported cases will be collected from all data sources. This indicator contributes to global AEFI reporting as part of the Global Vaccine Action Plan.
Source: WHO Global Advisory Committee on Vaccine Safety – Indicators
• Proportion of community-reported AEFI cases: This will be calculated as the ratio of AEFI cases reported by the community to the total AEFI cases reported. This indicator will track the percentage of total AEFI reports that come directly via SMS or community alerts compared to the overall AEFI cases reported.

3. For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?

Yes, the AEFI case definition should be temporarily broadened to maximize early signal detection for a newly introduced vaccine since it described that the AEFI surveillance system in Northern Nigeria is recognized as not robust enough to generate sufficient and convincing vaccine safety data, especially for new vaccines and those under emergency authorization use.

The trigger to revert to the prior definition would be used once the National Expert Committee confirms that the AEFI surveillance system has successfully generated sufficient vaccine safety data. This means the data must be robust enough to support accurate causality assessments with consistent reporting.

[Hteik Htar Tin]

Thanks for your discussion for using AEFI indicators, Ama. I have no idea to use these indicators and just focused on data completeness and timeliness.

[Jenny Bituin]

Thank you for sharing. I agree, regular meetings and giving acknowledgment and recognition to health facilities that report AEFI cases in a timely and consistent manner might help in encouraging other healthcare facilities to improve their AEFI case reporting.

[Than Htike Aung]

1.) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
The most significant design limitation that threatens valid estimates of both sensitivity and representativeness in the AEFI surveillance system is the exclusion of private health facilities and some general hospitals from the reporting network. Although surveillance system deploys in Routine Immunization (RI) place, AEFI can be happened anytime after vaccination and omitting some facilities leads to systematic under-reporting and biased estimates. This results in missed AEFI cases, especially from underserved or rural areas, ultimately compromising the system’s ability to detect all true cases (sensitivity) and to reflect the true distribution of AEFI occurrences in the population (representativeness).
To address this limitation within six weeks, the surveillance team could rapidly expand the inclusion of private and all public healthcare facilities into the AEFI reporting network. Short virtual or on-site training sessions could be organized for these facilities to orient staff on AEFI case definitions, data tools, and reporting timelines. A simplified one-page weekly reporting form or a mobile entry system through DHIS2 could be introduced to make reporting easier. In addition, surveillance officers should conduct active follow-up to ensure that each facility submits at least one report within the first month.
By implementing these targeted actions within a six-week period, the surveillance system would achieve broader coverage, reduce under-reporting, and significantly improve both sensitivity and representativeness, leading to more reliable and valid estimates of AEFI occurrence.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
A low-cost intervention to increase the sensitivity of the AEFI surveillance system, using the CDC surveillance attributes, would be to implement a monthly SMS reminder and feedback system for frontline health workers. This system would send simple text prompts reminding vaccinators to report any observed AEFI cases, including “zero reports” when no cases occur. In return, the system would provide short feedback messages acknowledging receipt of reports and sharing quick data summaries or recognition for timely submissions.

This intervention directly targets the sensitivity attribute by increasing case detection and reporting completeness through frequent engagement and accountability. It reinforces awareness of AEFI case definitions and ensures that even minor or delayed cases are not overlooked due to forgetfulness or competing work priorities.

The expected trade-offs include a potential increase in false-positive or minor AEFI reports, as more health workers may report mild or unrelated events. This may temporarily increase data volume and workload for data validators at the LGA or state level. However, these trade-offs are acceptable because higher sensitivity strengthens early warning capacity and public health responsiveness.

To measure the impact of the intervention, the following indicators could be tracked:

1. Percentage increase in reported AEFI cases per month after implementation (compared with baseline).
2. Proportion of health facilities submitting any AEFI report (including zero reports) each month.
3. Timeliness of AEFI report submission with proportion of reports received within the expected reporting timeframe.
Overall, this low-cost, communication-based strategy would enhance the system’s sensitivity by improving active participation, reducing under-reporting, and fostering a culture of consistent AEFI surveillance among health workers.

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?

Yes. For a newly introduced vaccine, it is appropriate to temporarily broaden the AEFI case definition to maximize early signal detection. During the initial rollout, there is limited real-world data on the vaccine’s safety profile, so widening the case definition increases the sensitivity of the surveillance system, helping to capture even rare or unexpected adverse events that might otherwise be missed. This proactive approach ensures that potential safety concerns are identified and investigated early, building public trust and allowing timely corrective actions if needed.

However, a broadened definition should be time-limited to prevent unnecessary over-reporting and data burden. The system should revert to the standard, more specific definition once certain triggers are met.By applying this temporary, sensitive approach early and returning to the routine definition once sufficient safety data are available, the surveillance system balances vigilance with efficiency, ensuring both early signal detection and long-term sustainability.

[Sirithep Pl]

1.) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?

The main limitation was reliance on passive, facility-based reporting with incomplete facility coverage and under-participation, which makes the system insensitive and non-representative.
To improve this limitation within 6 weeks, producing reporting at sentinel sites may be required. The plan could be addressed as follows:
– Preparation and training: preparation of tools (simple and feasible for usage) and training personnel with training material to understand and have an insight for surveillance.
– Surveillance: prospective active finding at sentinel and community with daily reporting of any AEFI and weekly home-based checks for a sample of recently vaccinated children/adults. Ensuring private facilities in each sentinel are enrolled as reporting sites and asked to send daily/weekly counts.
– Data cleaning & rapid capture–recapture estimate: producing a short report with estimated sensitivity and representativeness gaps and recommended scale-up.
– Feedback and adjustment: sharing results and using findings to expand the sentinel approach to more sentinels; immediately include private providers into routine reporting; update local communication to caregivers to report AEFI.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
Intervention:
– Targeted sentinel active surveillance using rapid reporting with SMS or simple application (e.g. WhatsApp, LINE) for daily check-in from a selected network of sentinel. Uses existing tools requires only modest spending, brief training, and simple line lists. This makes reporting active rather than only passive, captures minor events that otherwise are not reported, and ensures private clinics and rural communities are included at sentinel level.

Expected trade-offs / downsides:
– More false positives: broad catching will raise reports of events unrelated to vaccination (lower specificity and more investigations).
– Increased workload for triage/investigation — could overwhelm limited investigation capacity.
– Short-term resource needs: budgets, airtime, and coordination time.

Indicators to detect impact
– AEFI reporting rate and proportion should rise if sensitivity improves and shows engagement/coverage increase.
– Completeness of key fields on AEFI line lists improved data quality suggests usable additional sensitivity.

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?

Yes — A broader case definition (more sensitive, less specific) increases the chance of detecting rare or unexpected early safety signals during rollout that critical when background safety profiles are uncertain.

[Hteik Htar Tin]

Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
In this AEFI surveillance system, there is parallel different reporting systems. These two have no interoperability, usability and used different tools for data management.
For addressing this issue, I will start the re-designing of surveillance system reporting flow. To initiate I will do advocacy meeting with M&E team and Disease Surveillance Unit for the health staff’s awareness of the need for change, targeting to raise their desire to use acceptability interoperable data and data collecting tools (single one). Then, after getting consensus to use single management system, I will prepare the electronic data reporting system for AEFI cases from health facility to common server as immediate reporting. For zero reporting cases, there will be defined specific date (every Tuesday) to report to server. For stability strengthening, I will recruit new staff for surveillance data reporting, prepare the report formats as simple and flexible ones. After this, the AEFI surveillance training will be given to all RI providers, LIO, RIO, DSNO and M&E Officers, emphasizing the importance of data to report and giving constructive feedbacks and required logistics supplies to surveillance site.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
In this case, under-reporting and incomplete data decrease the sensitivity of system. So, I will initiate the direct electronic reporting system to central server (KoBo application can be used freely). All LIO, RIO, DSNO and M&E Officer will be grouped as surveillance team and monitor the reported data as their assigned duty. The reporting format will be prepared to collect the most required data (date, health facility name, village/ward, age, received vaccine, symptoms). After confirmation of the case, the case investigation process will be proceeded by surveillance team to reported health facility.
Expected trade-offs:
(1) Required training to use KoBo application and create KoBo server
(2) Responsiveness from the surveillance team will be faster and start investigation or feedback to health facility
(3) Knowledge and perceptions of RI providers to collect AEFI data will be raised due to proper training.
(4) Their workload can reduce compared to previous system.
(5) Data security and privacy must be applied for the system.
Indicators
(1) Timeliness and completeness of surveillance data on weekly basis
(2) Random data quality assessment reports between primary data source and electronic system reporting
(3) Feedbacks evaluation from RI providers for using electronic reporting system

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?
Yes, AEFI case definition should be temporarily broadened to get many positive case as much as. Because we have no adequate capacity in giving information about vaccine safety, no reported AEFI data means that the system is not competent to provide vaccine safety consideration.
To revert the prior definition, I will request for ad-hoc NEC meeting, reporting the system’s pitfalls to consider vaccine safety and modify the AEFI case definition for new vaccine to AEFI National Expert Committee (NEC).

[Jenny Bituin]

Thank you for sharing. Your solution to address the parallel reporting systems with no interoperability is very comprehensive.

[Siriluk Dungdawadueng]

Omoleke and de Kiev’s evaluation revealed that the AEFI surveillance system in Kebbi State had major weaknesses affecting sensitivity and representativeness. From a public health perspective, I think the most serious design limitation is under-reporting due to passive surveillance and limited stakeholder participation. Many healthcare workers may not recognize or report all AEFI cases, especially from peripheral or private facilities.
To address this within six weeks, I would introduce targeted active case finding in a few sentinel sites combined with short refresher trainings and supervision visits. This would quickly raise awareness, improve reporting completeness, and provide data to benchmark sensitivity.

1. Key design limitation and short-term action

Limitation: Reliance on passive reporting → high risk of missed AEFI cases.

Action: Conduct a six-week rapid active surveillance at selected high-volume sites and cross-check facility records against AEFI reports.

Rationale: This can estimate under-reporting and improve representativeness without redesigning the whole system.

2. Low-cost intervention to increase sensitivity

Using CDC surveillance attributes, one low-cost intervention could be:

Intervention: Implement a simple WhatsApp-based AEFI notification group for frontline health workers.

Expected trade-offs:

↑ Sensitivity and timeliness

↓ Data completeness and risk of unverified reports (false positives)

Indicators to track impact:

Number of AEFI reports submitted per month (baseline vs post-intervention)

Proportion of reports verified within 72 hours

Number of facilities actively contributing reports

This approach builds on simplicity and communication—two attributes already identified as partial strengths.

3. Broadening case definition for new vaccine

In my view, temporarily broadening the case definition makes sense at the start of a new vaccine introduction because early detection of unexpected signals is crucial.

Yes: It should be broadened initially to capture any unusual reactions.

Trigger to revert: After about two to three months or once the background AEFI rate stabilizes and no new safety signals emerge, the definition can revert to the standard one to avoid overburdening the system.

If the system capacity is already low, over-reporting could create false alarms—but early vigilance outweighs that short-term burden.

[Myo Thiha]

1) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?

In my opinion, the single design limitation most threatens valid estimates of sensitivity and representativeness would be personal bias of routine immunization (RI) providers who reported to the next level and not all facilities are conducting routine immunization, not all facilities are reporting, and not all cases are captured in the documentation. I would like to address it by mapping, enhancing standardization, providing training, implementing active reporting, monitoring, supervision and data quality checks, analyzing and feedback within 6 weeks.

2) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs and list 2–3 indicators to detect impact from the intervention.

I would like to propose SMS reporting system from caregivers and CHWs to report AEFIs in real time. Automated dashboard developed and monitored and conducted follow up by RI officer or LGA officer. The expected trade-offs are false positives and data noise.

The indicators are as follows:
i) AEFI reporting rate per 10,000 doses
ii) Proportion of reports investigated within 48 hours
iii) Proportion of ineligible case

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?

Yes, the AEFI case definition should be temporarily broadened to maximize early signal detection for a newly introduced vaccine. However, I would trigger to revert to the prior definition based on time, workload, and positive predictive value.

[Jenny Bituin]

1. Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
The stability or resilience of the AEFI surveillance system of the Kebbi State most threatens the valid estimates of sensitivity and representativeness. According to the study, 88.7% of respondents reported that lack of resources interrupted the AEFI surveillance system. This may lead to high degree of underreporting of AEFI cases, affecting the sensitivity and representativeness of the system. This can be addressed by increasing the investment to the lacking resources (financial resources, human resources, and logistics such as data tools) in order for the AEFI system to be fully operational.

2. Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
Since interviews and FGDs revealed that some health workers lacks positive attitude towards AEFI reporting and the community have low awareness about AEFI and its reporting, a low-cost intervention to increase the sensitivity of the system is by educating the health workers and the community about the importance of AEFI reporting and surveillance. The expected trade-offs are it will take some time to educate and convince everyone to report AEFI cases and will be added workload to the health worker tasked in educating others. Two indicators to detect the impact from the intervention are increased number of AEFI cases reported and increased participation of the community in AEFI reporting.

3.For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
Because it is a new vaccine, yes, the AEFI case definition should be temporarily broadened to maximize early signal detection. The trigger to revert to the prior definition is when high number of false positive cases were detected.

[Nang Phyoe Thiri]

Yes Jen, I agree with you. User’s attitude toward the system really affect whether the system is operational or not.

[Wai Phyo Aung]

1) ) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
Single design limitation threatens the validity of sensitivity and representatives. In the articles, the main problem is under-reporting of AEFI cases. It can affect the sensitivity for being delaying of information, incomplete data. Regarding to representatives, data could not represent the whole cases because of under-reporting and may interpret as false case load. In impact, it might mis-regard as not an important issue.

In the next six month, we can fix to improve real time reporting by establishing the electronic reporting system and can train the health staff to use the mobile reporting system. Moreover, we can train the health staff the technical competency to investigate and identify the AEFI cases in immunization.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
Regarding to low cost intervention;
Expected trade off; we can consider and ensure the system as the following;
1) Simplicity: reporting, data collection format should be simple and understandable for health staff. It should not be extra workload for them
2) Timeliness: Real time reporting is critical for AEFI cases. The defined channel must be reflected on ground situation. Delay report can not be used to develop effective response.
3) Data quality; The reported data must be fulfilled with standard data quality. Incomplete and missing data may take time and cause backward to interpret the cases.
we can set integrated reported indicator as 1) number of AEFI case in your health facilities; 2) Number of notified AEFI cases within 24hr in the monthly or quarterly report of Health information management system (HMIS) routine report.

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?

It is absolutely needed to broad the early detection when newly vaccine is released in the community. It can be reverted when sensitivity and Predict value positive rate are dramatically decrease after analyzing the system, root causes, workload, data quality.

[Salin Sirinam]

1.) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
– It was the lack of involvement of private healthcare providers, resulting in under-reporting and incomplete documentation. When this data was missed, it means the community awareness of AEFI report would be low. Therefore, I would suggest including them by implementing a fast-track training for staff in these centers. For example, scheduling the workshop for the key staff responsible for detecting and reporting AEFI. Then, providing them the necessary tools for detection and documentation. Direct consultation should also be offered to support the fast track process.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
– Since the lack of participation and low awareness was mentioned, a feedback loop could be implemented. It can be a regular monthly requirement for health center leaders to submit their reports to the central database. A simple summary from the central system should be produced and distributed so that the stakeholders can see their contribution, and also can help monitor vaccine safety data. The trade-off is this practice can add more workload to the local staff as well.
The key indicators could be: the proportion of complete reporting, and the timeliness of reporting.

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– Yes, which means that more inclusive case definition will increase sensitivity, which is important for a new vaccine. The trigger could be for example the standardized reporting rate is achieved, and the decision would be made baked on the judgement of the National Committee.

[Nang Phyoe Thiri]

Hi Salin. I agree that regular feedback loops and continuous coaching is mandatory for a system setup.

[Soe Wai Yan]

1. Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
The most critical design limitation is the system’s reliance on passive surveillance which threatens both sensitivity and representativeness. Health workers may lack training, motivation or time to report every AEFI and some communities may be completely missed.
To address this within six weeks, I would implement a short-term active surveillance pilot in a selection of underrepresented LGAs. This would involve:
a. Assigning trained health workers or community informants to actively seek and report AEFIs weekly,
b. Leveraging existing immunization outreach structures to minimize cost and delay,
c. Comparing reported cases from active sites to passive ones to estimate under-reporting.

2. Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
Intervention:
Distribute simple AEFI reporting job aids and provide short refresher training for frontline health workers during routine supervision visits targeting sensitivity.
Expected Trade-Offs:
May lead to an initial increase in reports of mild or unrelated events, reducing specificity that could temporarily burden health workers with additional tasks.
Indicators to Track Impact:
a. AEFI reporting rate per 10,000 vaccine doses that should increase if possible.
b. Proportion of health facilities submitting at least one AEFI report per month.
c. Completeness of AEFI report forms to ensure quality is maintained.

3. For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
Yes, temporarily broadening the AEFI case definition is appropriate during the early phase of a new vaccine rollout. It increases the likelihood of detecting rare or unexpected adverse events which is critical for public trust and early risk management.

[Nang Phyoe Thiri]

1.) Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?

Relying only on passive facility-based reporting threatens sensitivity and representativeness.
Within 6 weeks,
My plan is to combine passive and active surveillance methods.
Firstly, I will select sentinel sites for passive and active surveillance.
Train focal persons the reporting procedure for AEFI (the timing, interval, variable, channels) and close follow-up for support.
Additionally, the focal persons will conduct phone call follow-ups to caregivers of children after 7 days of vaccination to capture unreported events.
I will follow up with the focal persons about the reporting flow/active tracing flow and adjust when necessary.
I will review weekly data and conduct evaluation in 6 weeks.

2.) Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.

Low cost intervention – 7-day post vaccination follow ups with phone call at sentinel sites

CDC attributes addressed:
Sensitivity – capture more AEFI cases through active follow-ups
Timeliness- timely detection of any adverse effect
Acceptability – simple methods, easy to use

Trade-offs- More workload and burden to staff, phone bill costs, more false positives
Indicators –
1.AEFI reports/ 100,000 doses
2. % of AEFI cases detected through active phone call follow-ups
3. % of serious events detected within 48 hr

3.) For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?

Yes, the AEFI case definition should be temporarily broadened to maximize early signal detection.

When to revert – after 3 months and there is no new safety signal detected

[Kevin Zam]

1. Which single design limitation most threatens valid estimates of sensitivity and representativeness? How would you address it within six weeks?
The biggest problem is that the AEFI system depends mostly on passive facility reporting, which misses many cases in the community and private clinics. This affects sensitivity and representativeness.
How to address it in six weeks:
Train community health workers (CHWs) and private clinics to report AEFIs using WhatsApp, SMS, or a simple online form. Collect reports from at least two sources (e.g., clinics and CHWs) and compare them using a simple capture–recapture method to estimate missing cases. Combine all data in one Excel or Google Sheet and analyze weekly.

2. Using the CDC surveillance attributes, propose one low-cost intervention to increase sensitivity. State the expected trade-offs, and list 2–3 indicators to detect impact from the intervention.
Intervention: Let CHWs send reports of any health problems after vaccination through SMS or WhatsApp.
Trade-offs: We will get more reports, but many might not be true AEFIs (more work to verify).
Indicators to measure impact:
Number of AEFIs reported per 10,000 doses (should increase).
Median time from event to report (should decrease).
Percentage of reports verified within 48 hours (shows improved response).

3. For a newly introduced vaccine, should the AEFI case definition be temporarily broadened to maximize early signal detection?
– If yes, what trigger would you use to revert to the prior definition?
– If no, why should this change not be implemented?
Yes, the AEFI case definition can be temporarily broadened when a new vaccine is introduced. This helps catch more possible reactions early.
When to return to the old definition: After 8–12 weeks, or when confirmed serious AEFIs stay stable for a month, and the verification workload becomes manageable.
If not broadened: The system may miss early safety signals. However, if the team cannot handle many reports, it’s better to strengthen capacity first before broadening.

[Wah Wah Lwin]

It’s interesting that you choose the indicator “Median time from event to report” 🙂

[Author]

Posts