Strengthening timely detection and reporting of unusual respiratory events from health facilities in Yaoundé, Cameroon

Abstract Background The International Health Regulations state that early detection and immediate reporting of unusual health events is important for early warning and response systems. Objective To describe a pilot surveillance program established in health facilities in Yaoundé, Cameroon in 2017 which aimed to enable detection and reporting of public health events. Methods Cameroon’s Ministry of Health, in partnership with the US Centers for Disease Control and Prevention, Cameroon Pasteur Center, and National Public Health Laboratory, implemented event‐based surveillance (EBS) in nine Yaoundé health facilities. Four signals were defined that could indicate possible public health events, and a reporting, triage, and verification system was established among partner organizations. A pre‐defined laboratory algorithm was defined, and a series of workshops trained health facilities, laboratory, and public health staff for surveillance implementation. Results From May 2017 to January 2018, 30 signals were detected, corresponding to 15 unusual respiratory events. All health facilities reported a signal at least once, and more than three‐quarters of health facilities reported ≥2 times. Among specimens tested, the pathogens detected included Klebsiella pneumoniae, Moraxella catarrhalis, Streptococcus pneumoniae, Haemophilus influenza, Staphylococcus aureus, Pneumocystis jiroveci, influenza A (H1N1) virus, rhinovirus, and adenovirus. Conclusions The events detected in this pilot were caused by routine respiratory bacteria and viruses, and no novel influenza viruses or other emerging respiratory threats were identified. The surveillance system, however, strengthened relationships and communication linkages between health facilities and public health authorities. Astute clinicians can play a critical role in early detection and EBS is one approach that may enable reporting of emerging outbreaks and public health events.


| INTRODUC TI ON
In May 2016, an outbreak of avian influenza A (H5N1) virus "H5N1" occurred among poultry in Yaoundé, the capital city of Cameroon.
The outbreak caused widespread poultry mortality, required depopulation on select poultry farms, and resulted in the death of >15 000 birds. 1 The die-off was detected by Cameroon's livestock services.
Cameroon's Ministry of Health (MOH) was rapidly notified and within 24 hours activated their public health emergency operations center for rapid investigation and response for potential infection in humans. 2  Organization (WHO) standard case definitions for severe acute respiratory infection (SARI) and/or influenza-like illness (ILI) to detect possible influenza cases. [4][5][6] While the H5N1 outbreak illustrated rapid notification from the animal to human health sectors, the situation raised important questions about detection and reporting of potential human infections with avian influenza from hospitals. The existing IDSR and influenza surveillance systems in theory contributed to routine counting of SARI, ILI, and atypical respiratory cases; however, neither system reliably notified the MOH of suspected outbreaks in a timely way, despite specific instructions to do so within IDSR. During the H5N1 outbreak, it was unclear whether Yaoundé health facilities would recognize potential human infections with avian influenza, and if they did, whether the MOH would be alerted rapidly to initiate control measures. To improve timely detection and immediate notification of suspected outbreaks, the MOH, with support from the United States Centers for Disease Control and Prevention (CDC), implemented an event-based surveillance (EBS) program focused on detection of unusual respiratory events in Yaoundé.
Event-based surveillance systems, whether web or media based, focused at the community or healthcare facility level, are characterized by early detection and immediate reporting of potential public health events. They are seldom disease or pathogen-specific, and instead rely on pattern recognition. 7 While much of the EBS literature focuses on web or media EBS, [8][9][10] and community EBS, 8,[11][12][13][14] few reports describe the implementation of EBS in healthcare facilities. 13,14 The EBS program in Yaoundé was a collaboration among DLMEP, CPC, Cameroon's National Public Health Laboratory (Laboratoire National de Santé Publique de Cameroun, LNSP), and CDC. This report describes the implementation process, signal detection data, and the strengths and challenges of the program.

| ME THODS
Within the context of EBS generally, the WHO defines a signal as any data or information that could represent a potential acute risk to human health. 7 Each reported signal undergoes a process of triage and verification in order to ensure that a true public health event is occurring before public health authorities are activated. 7 The data sources that contribute to an EBS system (web, media, community, healthcare facility, etc) will influence the design of the data collection process and how the system will be tailored for its intended audience. 7 Astute clinicians can play a critical role in early detection and EBS is one approach that may enable reporting of emerging outbreaks and public health events.  Within the EBS framework, when a clinician or nurse detected a signal, they were instructed to notify the healthcare facility EBS focal point, who would immediately call CPC to report it ( Figure 1).
When CPC received the telephone call, they conducted the process of triage while on the phone, confirming that one or more of the four signals were met and that the current signal did not represent a duplicate of a previously reported signal (ie, that it was a true signal No ethical approval was required because the data used in this manuscript were from public health surveillance.

| RE SULTS
From May 2017 to January 2018, 30 signals were detected and reported ( Table 2). All healthcare facilities reported at least one signal, and more than three-quarters of the facilities reported two or more signals. Signal four, "any unusual cases of severe acute-onset respiratory illness" was reported 25 times (83.3% of all signals), and it was routinely reported throughout the 9-month pilot. The signal corresponding to a cluster of patients with respiratory disease was reported three times (10%), and the signal corresponding to healthcare worker illness was reported twice (6.7%). During the triage process, nine of the 30 signals (30%) were determined to be false and 21 (70%) were considered true signals that warranted collection of additional information. The signals that were determined to be false were so categorized because they did not match signal criteria; no duplicate signals were reported.
Following the verification process, 15

| D ISCUSS I ON
The H5N1 outbreak that occurred among poultry in 2016 resulted in rapid notification from animal to human health sectors, demonstrating strong one health multi-sectoral collaboration for avian influenza, one of the country's prioritized zoonotic diseases. 15 Yet, there was concern that rapid reporting from healthcare facility to public Similar to EBS in other countries and settings, the signals in this system were defined to enable healthcare workers to report when they identify unusual occurrences or patterns. 14 Astute clinicians can play a critical role in early detection and reporting of emerging outbreaks and public health events. 16  The events detected in this pilot were caused by routine respiratory bacteria and viruses, and no novel influenza viruses or other emerging respiratory threats were identified. DLMEP did not consider any of the events to warrant additional public health action.
Despite this, however, an important outcome of the pilot was that EBS helped promote behavior change among clinicians and health facility staff toward detection and reporting of potential outbreaks.
An example of this was at health facility E, where a suspected hospital-acquired infection was recognized and reported as a potential public health threat within 24 hours of detection. The EBS system enabled the clinician to recognize that a potential hospital-acquired infection could be a risk to public health and that notification was warranted and provided a mechanism to report to public health offi-

| DISCL AIMER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.