Implementation of the 7-1-7 target for detection, notification, and response to public health threats in five countries: a retrospective, observational study

Summary Background Suboptimal detection and response to recent outbreaks, including COVID-19 and mpox (formerly known as monkeypox), have shown that the world is insufficiently prepared for public health threats. Routine monitoring of detection and response performance of health emergency systems through timeliness metrics has been proposed to evaluate and improve outbreak preparedness and contain health threats early. We implemented 7-1-7 to measure the timeliness of detection (target of ≤7 days from emergence), notification (target of ≤1 day from detection), and completion of seven early response actions (target of ≤7 days from notification), and we identified bottlenecks to and enablers of system performance. Methods In this retrospective, observational study, we conducted reviews of public health events in Brazil, Ethiopia, Liberia, Nigeria, and Uganda with staff from ministries of health and national public health institutes. For selected public health events occurring from Jan 1, 2018, to Dec 31, 2022, we calculated timeliness intervals for detection, notification, and early response actions, and synthesised identified bottlenecks and enablers. We mapped bottlenecks and enablers to Joint External Evaluation (second edition) indicators. Findings Of 41 public health events assessed, 22 (54%) met a target of 7 days to detect (median 6 days [range 0–157]), 29 (71%) met a target of 1 day to notify (0 days [0–24]), and 20 (49%) met a target of 7 days to complete all early response actions (8 days [0–72]). 11 (27%) events met the complete 7-1-7 target, with variation among event types. 25 (61%) of 41 bottlenecks to and 27 (51%) of 53 enablers of detection were at the health facility level, with delays to notification (14 [44%] of 32 bottlenecks) and response (22 [39%] of 56 bottlenecks) most often at an intermediate public health (ie, municipal, district, county, state, or province) level. Rapid resource mobilisation for responses (six [9%] of 65 enablers) from the national level enabled faster responses. Interpretation The 7-1-7 target is feasible to measure and to achieve, and assessment with this framework can identify areas for performance improvement and help prioritise national planning. Increased investments must be made at the health facility and intermediate public health levels for improved systems to detect, notify, and rapidly respond to emerging public health threats. Funding Bill & Melinda Gates Foundation.


Introduction
The COVID-19 pandemic showed the importance of systems to detect, notify, and respond to disease outbreaks early and effectively at local levels to avert health, economic, and social effects at national and global levels. The pandemic highlighted limitations of existing measures of such capabilities, 1,2 including the Joint External Evaluation (JEE) and Global Health Security Index, which use static measures of capacity rather than assessing how systems function in real-world conditions. The Independent Panel for Pandemic Preparedness and Response recommended "a fundamental reassessment which better aligns preparedness measurement with operational capacities in real-world stress situations, including the points at which coordination structures and decision-making may fail". 3 Similarly, the 2005 International Health Regulations (IHR) Review Committee's report highlighted the importance of early alert, notification, and response and recommended that WHO strengthen its tools and processes for assessing core capacities, including the use of functional assessments. 4 These recommendations underscore the importance of assessing the real-world performance of the complex systems required for detection and response.
Health emergencies are complex events, and their timely detection and response require capabilities at multiple levels: patients must have access to and seek care when ill, diseases must be recognised and then confirmed, and results must be made available at all levels of the health system to initiate a response. Results from process mapping have shown that a systems approach-reviewing the step-by-step performance of systems under real-world conditions-identified bottlenecks that, once addressed, accelerated disease control during previous outbreaks. 5 Increasing efforts have been made to use measures of real-world performance both to assess system functioning and to identify important gaps for performance improvement. Intra-action reviews and after-action reviews have been leveraged to identify performance gaps and inform remediation plans. [6][7][8] WHO's Thirteenth General Programme of Work (GPW 13) is measured using the Triple Billion impact framework, with a key tracer indicator of timely detection, notification, and response to calculate system performance in health emergencies. 9 7-1-7 has been proposed as a target for outbreak detection, notification, and early response, whereby every suspected outbreak is detected within 7 days of emergence and reported to public health authorities within 1 day of detection, and seven early response actions are completed within 7 days from reporting to public health authorities, indicating timely initiation of response. 10 The 7-1-7 target and its application integrate several successful aspects from previous work, including the use of timeliness metrics, 11 real-world events for learning and bottleneck analysis (from intra-action reviews and after-action reviews), application of a systems approach (from process mapping), and quantification of systems timeliness (from the GPW 13). 7-1-7 draws on existing timeliness metrics and simplifies their presentation for communication and advocacy, provides a framework for performance improvement, and supplements them with targets. Clear targets (eg, 90-90-90 for HIV) can create accountability frameworks and facilitate communication of challenges to key stakeholders who might lack visibility on the complex systems of health emergencies. 12 We assessed the feasibility and utility of the 7-1-7 approach to evaluate systems performance for detection, notification, and response in five countries. We summarised retrospective event data from these countries to evaluate historical performance and identify bottlenecks to inform performance improvement.

Study setting
For this retrospective, observational study, we partnered with national public health institutes in Ethiopia, Liberia, and Nigeria; the Ministry of Health in Uganda; and the Ministry of Health as well as municipal and state health departments in Brazil to conduct retrospective reviews of events that had occurred between Jan 1, 2018, and Dec 31, 2022. In alignment with WHO's methods for measurement of timely detection, notification, and response, 13 countries selected events meeting IHR criteria for serious public health events, as described in IHR Annex 2, 14 which include notifiable events under the IHR and events not formally notifiable to WHO but for which the number of cases, deaths, or both is large

Research in context
Evidence before this study We searched PubMed for articles published between database inception and June 1, 2022, which had, in the title or abstract, the terms "timeliness" and either "outbreak" or "epidemic", and identified 409 unique articles. We supplemented the scholarly literature review with a grey literature review for material published between Jan 1, 2000, and June 1, 2022, including normative and technical guidance from WHO. Only papers published in English were included. Some groups have proposed using timeliness metrics to monitor performance of disease surveillance systems. However, these metrics have not previously been connected to simple and clear targets for advocacy and communication or incorporated into a performance improvement framework. Although several groups have published timeliness results, an absence of standardised timeliness metric definitions prevents comparisons or pooling of data across studies. One study evaluated timeliness of 296 outbreaks in the WHO African region and found that timeliness of detection improved from 2017 to 2019. A study in Nigeria used timeliness metrics to show that an intervention-establishment of a revolving outbreak investigation fund-led to improvements in response timeliness. However, these studies did not present consolidated bottlenecks or enablers or describe a performance improvement framework.

Added value of this study
We used specific targets for timeliness of detection, notification, and early response actions indicating response initiation to allow clear communication and to identify bottlenecks and enablers, which were then aggregated and analysed to identify common themes. Through implementation in five countries, we found that the 7-1-7 target provides a systems framework through which countries were able to assess their epidemic preparedness capabilities at community, health facility, intermediate, and national levels. We also found that use of the 7-1-7 target complements existing preparedness measures by identifying bottlenecks and enablers of response that are not captured by the Joint External Evaluation tool, providing supplemental information that is useful for prioritising activities in national action plans for health security.

Implications of all the available evidence
Given the ability of 7-1-7 implementation to quantify the realworld performance of detection, notification, and response systems and identify bottlenecks and enablers for timely action, other ministries of health and national public health institutes might be interested in implementing 7-1-7 to evaluate and improve their epidemic preparedness capabilities.
for the given place, time, or population; the event has the potential for high public health impact; or external assistance is needed to detect, investigate, respond to, and control the current event or prevent new cases. Results were intended to be used to prioritise national planning. Thus, rather than selecting a random sample of events, countries used a purposive sampling approach to prioritise events for selection that represented their risk landscape with data available.

Data collection
Countries conducted desk reviews of selected events to complete structured templates and supplemented existing outbreak, situation, and intra-action and after-action reviews with inputs from subnational and national-level responders. Government officials responsible for the detection and response to these events identified milestone dates (date of emergence [t 0 ], date of detection [t d ], date of notification [t n ], and dates of each of seven early response actions [t r1 -7 ]), as well as bottlenecks (conditions that delayed actions) and enablers (conditions that facilitated prompt actions). Definitions of timeliness milestones are provided in the panel and appendix (p 2).
During 2021-22, government officials responsible for detection and response to health events in each country convened for a 1-day workshop to validate findings from the desk reviews and build consensus on priority bottleneck areas that would benefit from performance improvement. Because data were collected from existing response documentation and did not include personally identifiable information, the Resolve to Save Lives ethical review committee determined that this activity did not constitute human subject research. In addition, the ministry of health, national public health institutes, health department, or appropriate ethical bodies or institutional review boards in each country reviewed the protocols, made a non-human subject research determination, and gave permission to publish aggregate data.

Data analysis
We consolidated data for all selected events with complete information for dates of emergence, detection, and notification into Microsoft Excel. We used Stata (version 17.0) to synthesise median timeliness measures and calculate the proportion of events meeting targets for timely detection (t d -t 0 ≤7 days), notification (t n -t d ≤1 day), and early response. To identify bottlenecks and enablers, we used two measures and targets for effective early response: at least one of the seven response actions was initiated within 1 day of notification (min[t r1 -7 ] -t n ≤1 day); and all applicable early response actions indicating initiation were completed within 7 days of notification (max[t r1 -7 ] -t n ≤7 days). If data were missing for any of the seven early response actions, we used the latest available date in the timeline as the date of completion of early response actions. The full 7-1-7 target was met when an event did not exceed any of the three targets. The rationale for selection of these targets has been described previously. 10 We categorised events into six groups (foodborne or waterborne pathogen diseases, vaccine-preventable diseases, vector-borne diseases, viral haemorrhagic fevers, respiratory diseases, and other events). 15 Other events included disease transmission among animals and chemical poisoning in humans.

Analysis of bottlenecks and enablers
We generated descriptive codes using a grounded theory approach 16 based on 129 bottlenecks and 162 enablers identified (free text) by workshop participants. A primary round of coding generated 34 codes for bottlenecks and 32 codes for enablers, which were further collapsed into a final code bank 17  To understand the alignment between operational capabilities identified by participants and the relevant technical capacities in the JEE (second edition) tool, 18 we mapped each of the free text bottlenecks and enablers to JEE indicators and the JEE score level (from 1 indicating lowest capacity to 5 indicating highest capacity) required to reach the level of capacity described in the bottleneck or enabler. Two raters (AFB, CTL, HJM, FTL, RAT, JSD, or RLY) scored each bottleneck and enabler, with discordant results discussed and scored via group consensus. We used the JEE tool because it contains more indicators (n=49) than the State Party Self-Assessment Annual Reporting (SPAR; 2018 edition tool; n=24). 19 If insufficient detail was available to identify a corresponding JEE indicator, we coded the value as missing; if the bottleneck or enabler did not map to an existing indicator, it was coded as not applicable. If a bottleneck or enabler aligned with a capacity measured by a JEE indicator, but none of the qualitative attributes used to assign a score reflected the specific content of the bottleneck or enabler, we coded the score as not applicable. If attributes were relevant for two scoring levels, we coded the lower value for consistency.

Role of the funding source
The funder of the study had no role in study design; collection, analysis, and interpretation of data; writing of the report; or the decision to submit for publication.

Results
Of the 41 events assessed, 11 (27%) were viral haemorrhagic fever outbreaks (Rift Valley fever, Lassa fever, or Crimean-Congo haemorrhagic fever), ten (24%) were vaccine-preventable disease outbreaks (measles, polio, and yellow fever), six (15%) were respiratory disease outbreaks (COVID-19 or influenza A), five (12%) were vector-borne disease outbreaks (Rickettsial infection, Chagas disease, dengue, or chikungunya), five (12%) were foodborne or waterborne disease outbreaks (botulism, cholera, or other foodborne disease outbreak), and four (10%) were other events (disease in animals or chemical poisoning in humans;   Of 291 total bottlenecks and enablers observed, 248 (85%) could be crosswalked to existing JEE indicators, particularly those for reporting channels (45), case management procedures including health facility implementation of case definitions and standard operating procedures (31), availability of human resources (22), laboratory diagnostic capacity (17), and funding availability for timely response to health emergencies (16; table 5). Three (1%) bottlenecks or enablers did not have sufficient information to assign a JEE indicator, and 40 (12%) bottlenecks or enablers were not represented by JEE indicators, including access issues (conflict or remote settings; eight), COVID-19-related prioritisation challenges (eight), technological challenges including lack of mobile network coverage (seven), and low community knowledge or trust in the public health system (four).
Of the 248 bottlenecks and enablers that aligned with a JEE indicator, 40 (16%) could not be assigned a score because JEE scoring criteria did not capture the specified capability. Of the 208 bottlenecks and enablers that could be assigned a JEE score, 169 (81%) required a relatively high score (4 or 5) for sufficient capacity to achieve timely detection, notification, and response.

Discussion
In this retrospective review of public health events in five countries, we applied the 7-1-7 target to measure country capabilities for detection, notification, and early response initiation. We found that the median performance across all events suggests strong performance (6 days for detection, 0 days for notification, and 8 days to complete early response actions), but only a minority of events met all targets. Bottlenecks were most common at the subnational level, including the health facility level. Use of the 7-1-7 approach can supplement existing preparedness measures by identifying operational gaps from real-world events that might not be represented in the JEE (eg, community knowledge or trust in public health systems, and access issues) and can support prioritisation of national planning.
Timeliness metrics have previously been used to identify gaps and demonstrate improvements in timeliness of disease detection and reporting. 11,15,20 However, fewer studies report on timeliness of response to events, and there is no indication of progress on outbreak response and mitigation. Countries have used timeliness data to review disease surveillance system performance. A multi-country effort in the Mekong basin analysed data from 2087 outbreaks and found that dates of index onset, report, and response were more than 95% complete in all countries. 21 Similarly, a study from Nigeria used timeliness data (time to detection, notification, and response activities) to identify bottlenecks in detection and response to a Neisseria meningitidis serogroup C outbreak. 22 In this study, we describe the use of a simplified metric for reporting on timeliness that aligns with and supports implementation of the WHO Triple Billion impact framework. The 7-1-7 target was designed to align with and support implementation of the IHR, specifically capacities described in Annex 1   level, and national level. 14 The early response actions supplement the IHR Annex 1 to distinguish early response efforts from capacities required for extended responses for larger events. 10 Although we found that early responses were initiated within 1 day for most events (80%), fewer than half of events met the 7-day target for completion of all seven actions that comprise effective response initiation. Prompt response initiation was necessary but not sufficient for an effective early response. The 7-1-7 target highlights the importance of initiating a response coordinated across multiple pillars and can be used to assess the effectiveness of surveillance and response systems (eg, integrated disease surveillance and response in the WHO African region), as well as identify and advocate for implementation of appropriate performance improvement measures. For this analysis, countries conducted retrospective reviews of events that represented their risk landscapes. The approach to bottleneck and enabler identification and classification can be replicated in intra-action reviews, after-action reviews, and simulation exercises using the timeliness metrics described in WHO's country implementation guidance to review performance during the early phase of an outbreak or emergency and prioritise recommendations. 23 These metrics and bottlenecks can inform development and prioritisation of national planning, including national action plans for health security and multi-sector development plans. Because 7-1-7 evaluates the performance of systems involved in health threat detection and response and generates recommendations that should be translated into national planning and advocacy, its implementation is best domiciled in a relevant public health agency with the mandate for surveillance, response, and preparedness planning (eg, IHR national focal point).
Prospective implementation of 7-1-7 can reduce the retrospective data collection burden and guide realtime performance management for ongoing events. 5 Data collection should ideally be integrated into event management systems, 24 but can also be collected in response coordination tools or situation reports. Identification of bottlenecks and enablers can be best documented by the teams involved in the initial event investigation and response, either within rapid response team reports or through review of performance of an  ongoing or a recent event against the 7-1-7 target. Teams can mitigate desirability bias by adhering to 7-1-7 milestone definitions and using a self-assessment approach (similar to SPAR and after-action reviews) focusing on identifying bottlenecks to achieve performance improvement. Our analysis indicates that the 7-1-7 target is achievable across pathogen types and highlights the need for continued system strengthening. Most event types met at least two of the three targets, apart from vaccinepreventable diseases. Although substantial improvements have been made in surveillance and response for viral haemorrhagic fever outbreaks, functional capacities for surveillance need to be strengthened particularly for vaccine-preventable and vector-borne diseases, for which early responses can guide the targeted deployment of countermeasures and accelerate goals towards elimination or eradication. In addition, the finding that detection targets were not met for 33% of respiratory events (COVID-19 or influenza A) highlights that there are gaps in preparedness even for anticipated events and shows the value of the 7-1-7 target for both anticipated and unanticipated events. The event types for which notification and response delays were longest often required coordination of multiple sectors (eg, animalborne or vector-borne disease events) or multiple units within the ministry of health (eg, vaccine-preventable or foodborne or waterborne diseases). The findings highlight the importance of operationalising multi-sectoral coordination mechanisms for public health events that include routine communication channels for sharing epidemic intelligence and incident management protocols for multi-sectoral response governance.
Five pilot countries have used 7-1-7 to inform performance improvement efforts at national and subnational levels. Assessing actual performance during an event identified operational gaps that might not have  been identified or prioritised for improvement by existing tools and metrics and integrated priority actions into national operational plans. Their experiences show that investments in disease detection must increase at the health facility level, where most events are detected by clinicians outside the public health system. Clear communication and reporting channels between health workers and surveillance officers are crucial to verify events and initiate a larger public health response. We found that response bottlenecks most frequently involved resource limitations, including flexible funds for deployment of teams and availability of countermeasures, at the subnational level. National level resource availability to augment these gaps was a frequent enabler. In Nigeria, establishment of a flexible and rapid funding mechanism for early outbreak investigation decreased the median time to respond from 6 days to 2 days. 25 Implementation of 7-1-7 by countries at national and intermediate public health levels might similarly identify catalytic investments to detect and respond to public health events more quickly.
Our approach has limitations. First, 7-1-7 measures a subset of systems required for preparedness-those required for early detection and early response rather than for later-stage or larger-scale responses-and does not include the development, production, or ability to scale implementation of countermeasures or to prevent the initial emergence of dangerous pathogens. Second, our retrospective review methodology was restricted by historic data availability, limiting the number of assessed events, and self-assessment might have introduced desirability bias. The non-random sample of events prohibits further statistical inference, although we found that descriptive analyses identified common themes that may suggest potential root causes. Last, the countries involved in piloting the 7-1-7 approach are not a representative global sample and we are unable to generalise their performance or their observed bottlenecks.
As next steps, we propose and have initiated research in three areas for future evaluation. First, several more countries have initiated prospective implementation of 7-1-7 to capture data on all serious public health events, which will create a more representative dataset and generate a larger sample of events that would allow for disaggregated analyses of correlates of 7-1-7 performance by pathogen type and transmission scenario, improving our understanding of representativeness and generalisability. Second, we will evaluate 7-1-7 performance characteristics by assessing internal validity and interrater reliability of timeliness measures. Last, we propose an evaluation of associations between timeliness and impact metrics, such as morbidity and mortality, to assess external validity of the target.
We have found value in implementing 7-1-7 as both a performance metric and a systems tool for performance improvement. We recommend its adoption for real-time monitoring and performance evaluation during public health events, with routine synthesis of data to track progress on the functioning of systems required for early detection and action and to initiate and target rapid quality improvement processes. Adopting 7-1-7 as a monitoring and evaluation tool at national level can help countries to prioritise investments and capacity-building actions and to support measurement of global progress through the GPW 13 and regional integrated surveillance strategies. We have found that 7-1-7 implementation can also engage stakeholders outside of the health emergencies sector and identify clear targets for joint decision making and advocacy for more rapid detection and response to health threats.

Contributors
AFB, AM, and CTL conceived of the evaluation, with RLY, JSD, FTL, and HJM contributing to the evaluation design. IM, OA, AA, RJ, RLY, JSD, TOY, LN, AKY, RAT, ML, IN, JO, IMOA, and WO supported acquiring, managing, and interpreting the data, and AFB and CTL accessed and verified the data. AFB, RLY, JSD, FTL, HJM, RAT, and CTL supported the analysis. The manuscript was prepared by AFB and CTL, and all authors contributed to revision of the manuscript. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Declaration of interests
We declare no competing interests.

Data sharing
Data requests can be channelled to each participating country by the corresponding author (CTL).