Case study: design and implementation of training for scientists deploying to Ebola diagnostic field laboratories in Sierra Leone: October 2014 to February 2016
Abstract
As part of the UK response to the 2013–2016 Ebola virus disease (EVD) epidemic in West Africa, Public Health England (PHE) were tasked with establishing three field Ebola virus (EBOV) diagnostic laboratories in Sierra Leone by the UK Department for International Development (DFID). These provided diagnostic support to the Ebola Treatment Centre (ETC) facilities located in Kerry Town, Makeni and Port Loko. The Novel and Dangerous Pathogens (NADP) Training group at PHE, Porton Down, designed and implemented a pre-deployment Ebola diagnostic laboratory training programme for UK volunteer scientists being deployed to the PHE EVD laboratories. Here, we describe the training, workflow and capabilities of these field laboratories for use in response to disease epidemics and in epidemiological surveillance. We discuss the training outcomes, the laboratory outputs, lessons learned and the legacy value of the support provided. We hope this information will assist in the recruitment and training of staff for future responses and in the design and implementation of rapid deployment diagnostic field laboratories for future outbreaks of high consequence pathogens.
This article is part of the themed issue ‘The 2013–2016 West African Ebola epidemic: data, decision-making and disease control’.
1. The 2013–2016 Ebola virus epidemic in West Africa
The first recorded case of Ebola virus disease (EVD) occurred 40 years ago in September 1976 in Yambuku, a small village in then Zaire (now Democratic Republic of Congo, DRC). Up until December 2013, 24 reported Ebola virus (EBOV) epidemics affecting humans, involving 2388 infected persons and 1590 fatalities had been reported [1]. The West African outbreak of 2013–2016 is the largest recorded, with 11 310 fatalities and 28 616 infected persons as of the final WHO Ebola situation report of 10 June 2016 [2]. EBOV is a member of the family Filoviridae; the species responsible for this epidemic and an isolated parallel epidemic in the DRC in August 2014 is EBOV Zaire [3]. The 2013–2016 EVD epidemic began in The Republic of Guinea in December 2013 and rapidly spread to the neighbouring countries of Sierra Leone and Liberia in early 2014. In March 2014, the World Health Organization (WHO), Médecins Sans Frontières (MSF) and the European Mobile Laboratory (EM Lab) provided patient care and timely diagnosis of patients [4]. It was not until 8 August 2014 that the WHO declared the situation as a Public Health Emergency of International Concern (PHEIC), which remained in place until 29 March 2016. In response to the WHO's call for international support, the UK government committed to a package of aid for Sierra Leone, funded through the UK Department for International Development (DFID). This included building three EVD treatment centres (ETCs) at Kerry Town, Makeni and Port Loko, each operated by different non-governmental organizations (NGOs). Collocated with each Ebola Treatment Centre (ETC) was an EVD diagnostic laboratory run by Public Health England (PHE) (figure 1).
Figure 1. Geographical location of PHE EVD Field laboratories and Legacy laboratories. *PHE Makeni was at the Makeni ETC until December 2015 but is now a purpose built Legacy laboratory at the Makeni Government Hospital. (Online version in colour.)
2. Field laboratories and their roles in the 2013–2016 Ebola epidemic in West Africa
Early diagnosis of infection and pathogen identification are paramount for patient care and outbreak control, and field diagnostic laboratories have previously been successfully operated during EBOV epidemics [5]. More recently, mobile laboratories with molecular testing capacity have been deployed in epidemics of other high-biosafety-level pathogens including Marburg virus and Lassa virus [6]. The EM Lab [4], the Centres for Disease Control and Prevention (CDC) and other members of the WHO Global Outbreak Alert and Response Network (GOARN) have experience in responding to such epidemics. At the peak of the epidemic, approximately 40 such field laboratories from several international agencies were deployed to West Africa. While several of these had experience in deploying to outbreaks, few had direct involvement in previous EVD outbreaks [7]. PHE's diagnostic experience in assay development in previous outbreaks and its significant involvement with the EM Lab consortium allowed for several molecular and serological assays to be available for consideration for use in the field. PHE opted to use real-time reverse transcription polymerase chain reaction (qRT-PCR) for the detection of EBOV genome and a lateral flow device for malaria diagnosis in patient samples. Additional evaluation of novel diagnostic assays, including antibody lateral flow devices to detect EBOV antigens in patient blood samples [8,9] automated nucleic acid extraction and RT-PCR systems took place [10,11]. Blood chemistry analysis platforms were implemented later in the epidemic for the management of patient care. These analyses contributed to the validation of novel diagnostic assays and informed regulatory decisions about the use of novel tests for EVD diagnosis during and following the epidemic.
3. Timeline of UK Ebola virus disease field laboratory implementation in Sierra Leone
On 24 September 2014, DFID formally requested support from PHE to design, equip and operate EVD testing laboratories in Sierra Leone with a view to the laboratories being sustainable beyond the immediate crisis. The design for the first laboratory at Kerry Town had to fit into a pre-determined building footprint. A mock laboratory, fitted out to the same dimensions and specifications as the Kerry Town laboratory, was constructed at PHE–Porton in early October and was used to prepare and train all volunteers deployed to Sierra Leone. The Novel and Dangerous Pathogens (NADP) Training group at PHE–Porton developed and delivered week-long training courses for all volunteers deploying to the PHE laboratories in Sierra Leone. The first course ran from 13–21 October 2014 and trained an initial team of 12 scientists who deployed to Sierra Leone to prepare the Kerry Town laboratory, which received its first samples on 27 October 2014. Within a 33 day period, the Kerry Town laboratory was built, assays and protocols developed, equipment sourced and delivered to Sierra Leone and volunteers recruited, trained and deployed. The Kerry Town laboratory supported an 80 bed ETC run by Save the Children and a 12 bed facility for health workers operated by the UK Ministry of Defence (figure 2). The laboratories in Port Loko and Makeni become operational on the 5 and 8 December 2014, respectively, closing in December 2015 when the ETCs were decommissioned. A legacy capability to test for a wide range of pathogens of public health concern, in support of the Sierra Leonean recovery plan and local capacity building, has been established and is discussed in §15.
Figure 2. DFID ETC in Kerry Town, Sierra Leone (October 2014). Source: Ricci Coughlan/DFID. (Online version in colour.)
4. Field laboratory design
The PHE field laboratories were designed for the diagnosis of EVD by qRT-PCR. They consisted of four distinct areas: sample reception (I), the flexible film isolators (FFI) (II), the main laboratory (III) and the molecular analysis suite (IV) consisting of three distinct molecular areas for mastermix preparation, template addition and RT-PCR, respectively (figure 3, I–IV). A unilateral direction of sample processing was implemented to reduce any contamination risks. Patient blood samples from the ETC, local holding centres and buccal swab samples from the community were received at the laboratories on a daily basis between 6.00 and 22.00. All samples were processed to the point of virus inactivation within custom designed FFI under negative air pressure (−30 to −60 Pascals) filtered through two sequential high efficiency particulate air (HEPA) filters. Each FFI was backed up by a car battery and invertor; in the event of a power disruption, each could continue to run at negative pressure for up to 8 hours. When using the FFI, operator personal protective equipment (PPE) consisted of a back-fastening laboratory gown, eye protection and a base pair of laboratory nitrile gloves. The FFI user placed their hands in the thick gloves attached to the sleeves of the FFI and donned an additional pair of nitrile gloves over those of the FFI. At all times when handing infectious material, the operator was therefore triple gloved and working under negative air pressure. A full-face visor and disposable apron were worn in addition to the normal laboratory PPE when working with 5000ppm chlorine outside of the FFI.
Figure 3. Internal layout of PHE laboratories in Port Loko and Makeni ETCs. Yellow circles, infectious waste bins; GX, GeneXpert platform; EZ1, automated nucleic acid extraction platform; blood chem, blood chemistry analysis platforms; green arrow, personnel entry to laboratory; red arrows, sample entry to the laboratory.
5. Laboratory workflow
The workflow is detailed in figure 4, with each numbered step (1–15) described below. Upon pre-examination of the samples for appropriate packaging, they were accepted into the laboratory (1), allocated a unique patient unique ID and entered onto a sample tracking form (2). The patient information was then uploaded onto the laboratory secure database using the unique ID as an identifier (3). Samples along with replicate patient ID labels were transferred into the flexible film isolator (FFI, 4), the only area of the laboratory in which infectious material was handled directly. Sample types predominantly consisted of buccal swabs in transport media from the community and whole blood (WB) samples from the ETC. They were removed from their secondary and primary containers and transferred to 1 ml aliquots in smaller (2 ml) primary containment o-ringed externally threaded tubes. A small volume of WB was used to test for malaria using a rapid test (5). This allowed for a malaria result to be reported to the ETC within 30 min of the test being run (8). WB aliquots were prepared for downstream blood chemistry analyses and the validation of novel diagnostic platforms introduced throughout the epidemic (10), remaining aliquots of samples were doubly contained and stored at −20°C (12). Plasma was used for nucleic acid extraction. 1 ml whole blood (WB) aliquots were centrifuged for 10 min at 3000 r.p.m. prior to plasma being transferred to new 1 ml tubes (7). Specific volumes of plasma (9) or buccal swab transport media were then chemically inactivated prior to being removed from the FFI after surface decontamination with 5000 ppm chlorine for the appropriate 10 min contact time (11). Total nucleic acid was then extracted by either using manual or automatic protocols (11). Resulting nucleic acid was stored at 4°C (12) until added in the template room to newly prepared qRT-PCR mastermix reaction tubes that were then placed on the Cepheid Smartcycler II Real-Time qRT-PCR platform (13). Sample results from all assays were then verified and cross-checked prior to being entered on a secure laboratory database (14). Finally, all data were revalidated by the laboratory team lead prior to being released and used to populate the daily situation reports (15). The complete process from sample reception to release of a qRT-PCR EBOV diagnosis took between 5 and 6 h.
Figure 4. Overview schematic of the workflow of the PHE EVD field laboratories in Sierra Leone.
6. Malaria rapid test on whole blood
Commercially available malaria rapid diagnostic tests (RDTs) were used within the FFI to initially screen patient WB for Plasmodium species responsible for causing malaria. Patient WB (15 µl) was placed onto an RDT cartridge, flushed through with appropriate assay buffer and a result recorded 15 min later. Based on the experiences of the EM Lab in Guinea, the BinaxNOW RDT [12] was used initially as it tested for four species of Plasmodium that infect humans (Plasmodium falciparum, P. vivax, P. ovale and P. malariae). However, as the epidemic continued, a call to standardize the malaria tests being utilized by all diagnostic laboratories within Sierra Leone led to the SD Bioline Malaria Ag P.f. RDT [13] being adopted, with improved specificity for sole detection of P. falciparum, the predominant Plasmodium species in Sierra Leone [14].
7. Plasma separation and sample inactivation
Optimization of the qRT-PCR assays and communication with the EM Lab demonstrated that extracting virus nucleic acid from separated plasma rather than WB resulted in a significant reduction in PCR inhibition. Within the FFI, two 0.8 ml volumes of patient WB were transferred into 2 ml o-ringed tubes and centrifuged for 10 min at 3000 r.p.m. The plasma was collected and pooled into a clean 2 ml o-ringed tube. A subset of the pooled plasma or swab collection media was inactivated in AVL lysis buffer containing the chaotropic agent guanidine thiocyanate (Qiagen GmbH, Germany) as follows. For automated nucleic acid extraction, either 80 µl plasma or 25 µl swab transport media plus 55 µl water was added to 320 µl AVL buffer. For manual nucleic acid extraction, either 50 µl plasma plus 90 µl water or 25 µl swab transport media plus 115 µl water were added to 560 µl AVL buffer. Sample/AVL mixtures were mixed, pulse centrifuged and a 10 min inactivation period was recorded. Manual extraction involved the additional inactivation step of adding 560 µl of 100% ethanol to the sample/AVL mixture, mixing, pulse centrifuging and leaving for an additional 5 min. Inactivated samples tubes were removed from the FFI after surface decontamination with 0.5% chlorine and a cessation of all FFI work for a contact time of 10 min. Tubes containing virus inactivated samples were then processed on the laboratory bench, following a pulse centrifugation. Manual RNA extraction proceeded according to the manufacturer's protocol (QIAamp Viral RNA mini, Qiagen GmbH, Germany). Remaining plasma was retained in 2 ml tubes that were surface decontaminated, double-bagged and removed from the FFI for storage at −20°C in the main laboratory following the default 10 min decontamination with 5000 ppm chlorine.
8. RNA extraction (automated and manual)
Automated extractions were carried out using an EZ1 nucleic acid extraction machine (Qiagen GmbH, Germany). EZ1 kits incorporated a multi-reagent strip that included ethanol, hence ethanol was not added during the inactivation process in the FFI, as was done for samples for manual RNA extraction. In order to ensure that samples received two inactivation treatments before being handled on the bench for automated extraction, tubes were pulse centrifuged, incubated for 15 min on a heat block at 60°C, and processed according to the manufacturer's protocol (QIAamp EZ1extraction, Qiagen GmbH, Germany). For both extraction methods, purified RNA including an internal MS2 control were eluted in 60 µl of AVE elution buffer and transferred to the refrigerator in the template room.
9. RT-PCR preparation
Laboratories initially operated the RealStar Filovirus RT-PCR kit 1,0 (Altona Diagnostics GmbH). In part due to operational requirements and in part due to performance issues, this was subsequently changed to a real-time RT-PCR assay based on the Zaire NP target assay [15] although modified to include a multiplexed MS2 internal control and to use a different reagent mix [11]. To limit on-site reagent preparation, custom Ebola Zaire and Internal control primers sets were supplied as pre-mixes at the required concentration by Thermo Fisher. Working stocks of the reaction mix were prepared in the dedicated mastermix room. These were then transferred to the template room where extracted sample nucleic acid was added prior to PCR, which was performed in a separate PCR room. The mastermix and template rooms were equipped with PCR cabinets that were cleaned with RNase Away (Molecular BioProducts, USA) between each use and also had UV facility to minimize risk of contamination.
10. Standard operating procedure development and assay optimization
Prior to laboratories commencing testing, all risk assessments (RAs), laboratory codes of practice (COP) and standard operating procedures (SOPs) for the diagnostic assays and equipment were developed at PHE. This was a dynamic process and involved several version updates in order to deal with the ever evolving situation on the ground in Sierra Leone. In-country feedback was provided by team leaders so that SOPs could be updated to ensure volunteers were consistently trained on the latest versions. Additionally, as the outbreak continued, additional diagnostic equipment for field assessment to support patient care were introduced, for which additional SOPs and training were implemented for pre-deployment training of volunteers.
11. Pre-deployment training for UK volunteers
A week-long pre-deployment training programme that included theoretical and practical aspects of working with EBOV and covered biosafety, biosecurity, emergency procedures, diagnostics and management of the laboratory units was designed by NADP Training. It was also aimed to psychologically prepare volunteers to deploy to a genuine outbreak situation. Information and feedback from NADP Training staff who had undertaken the EM Lab training [7] and had deployed to Guinea were used to help design the PHE training programme. A mock laboratory replicating the layout and dimensions of the Kerry Town laboratory was installed at PHE–Porton to accommodate training in the equipment and assays associated with the sample workflows (figure 5). Additionally, volunteers were trained to function as a team in realistic environmental working conditions with scenarios and challenges provided to mimic potential trials they might encounter in the field.
Figure 5. (a) PHE–Porton replica of the Kerry Town EVD laboratory (embedded image) prior to completion of refurbishment by The Royal Engineers. (b) Training and use of the flexible film isolators in the PHE–Porton training laboratory and Kerry Town ETC laboratory, respectively. (c) RNA template addition room at the PHE–Porton training laboratory and Makeni ETC laboratory, respectively.
12. Recruitment of volunteers
Requests for volunteers were sent to all PHE staff, NHS laboratories, UK universities and any other organizations that might provide suitable staff, such as the Animal and Plant Health Agency (APHA). As individuals volunteered with the NGOs rather than with PHE directly, their employment status did not change (they remained employed with their substantive employer and were released from their normal duties in order to volunteer). A direct agreement was established between the NGO running the ETC and the volunteer. Briefly, the selection process entailed volunteers contacting a dedicated email address for further information. They were then asked to complete a selection form, which requested details of their scientific experience and foreign travel/work, together with a brief statement of their reasons for volunteering. The selection criteria included: having knowledge of basic microbiology, being scientists or technicians with a background in infectious disease diagnostics, having recent laboratory experience, ideally including molecular experience. Attitude, ability to work well with others and cope in an unfamiliar environment and respond appropriately in an emergency were also deemed of high importance, with these criteria repeatedly being tested throughout the NADP training programme. A section for the volunteer's line manager to complete included questions on a range of topics such as the volunteer's ability to cope in a crisis. A small team of managers at PHE–Porton reviewed all the selection forms and assigned a traffic light rating on the volunteers' suitability for deployment. Those falling within the red rating were not invited to pursue their application, those in the amber rating were asked to attend the training course and invited to deploy based on assessment of their performance during the training course. Those falling into the green rating were invited to complete training. All volunteers were required to communicate with occupational health to assess if any underlying health problems were likely to be an issue if deployed and to address vaccination requirements. Each NGO also required volunteers to complete a psychological evaluation by phone and an online security training module.
Of the 402 volunteers trained between October 2014 and August 2016, we present data for 376 with complete demographic details including age, sex, employer and nationality (table 1). Of these, 39.6% were male and 60.4% female (table 1). The age groups of the volunteers ranged from the youngest aged 22, to the eldest aged 71 years old. Approximately half (48.1%) of all volunteers were aged between 26 and 35. Volunteers were recruited from within PHE, other government agencies, NHS laboratories and UK academic institutions; 78.2% were British, 17.3% were from the European Union and 4.5% from outside the UK and the EU. In total, volunteers from 28 countries were among those trained and deployed to PHE Laboratories in Sierra Leone.
age | total (%) | male (%) | female (%) | total (%) |
---|---|---|---|---|
22–25 | 30 (8.0) | 13 (43.3) | 17 (56.7) | 30 (100) |
26–35 | 181 (48.1) | 54 (29.8) | 127 (70.2) | 181 (100) |
36–45 | 80 (21.3) | 38 (47.5) | 42 (52.5) | 80 (100) |
46–55 | 57 (15.2) | 33 (57.9) | 24 (42.1) | 57 (100) |
56–65 | 25 (6.6) | 9 (36.0) | 16 (64.0) | 25 (100) |
66–71 | 3 (0.8) | 2 (66.7) | 1 (33.3) | 3 (100) |
22–71 | 376 (100) | 149 (39.6) | 227 (60.4) | 376 (100) |
13. Pre-deployment Ebola virus disease laboratory training programme
Training was designed to train individuals to meet SOPs and guidelines established by PHE for work with high consequence pathogens, Advisory Committee on Dangerous Pathogens (ACPD) guidance published literature and equipment manufacturers' manuals. Importantly, input from partner institutes' expertise and experience was included, in particular the EM Lab consortium, having already operated in Guinea and Liberia from March 2014.
Practical training was supervised and volunteers were assessed on whether they were proficient in the tasks they were expected to perform. A record of training was completed for each volunteer to include written assessment of the COP and SOPs. Each volunteer performed the procedures and also acted as a buddy for others for various steps, especially safety critical steps. Tasks were ranked into those that were essential for all to achieve competence. Volunteers recorded if they performed and buddied on each technique, and were signed off if they felt competent in the specific techniques. The trainers then signed the training record when satisfied with the level of competence achieved by an individual. If an individual had not been deemed competent on non-essential SOPs then they could only perform the tests that they had been signed off on, but received on-going support and training in-country on other areas until deemed competent (table 2).
day | training material covered in module | methodology | time (h) |
---|---|---|---|
1 | Introduction to Ebola and the W. African epidemic & biosafety | lecture | 1 |
1 | Risk assessment—field & laboratory | lecture | 1 |
1 | Psychology of working in W Africa during an Ebola epidemic | group discussion | 1 |
1 | Laboratory code of practice (quality and safety) | lecture | 1 |
1 | Laboratory overview—PPE and safe glove changes using UV dye | demonstration & practical | 1 |
1 | Laboratory overview—waste disposal & preparation of hypochlorite | demonstration | 1 |
2 | Sample reception & database entry | demonstration & practical | 2 |
2 | Overview of samples processing and GMP practical on bench | demonstration & practical | 1 |
2 | Flexible film isolator (FFI) overview, gauntlet change & set-up | demonstration & practical | 1 |
2 | Processing of mock samples in the FFI | practical | 3 |
2 | Disinfection, removal of samples from the FFI & heat inactivation | practical | 0.5 |
3 | Nucleic acid extraction—manual | practical | 1 |
3 | Nucleic acid extraction—automated (Qiagen EZ1) | practical | 1 |
3 | Real-time RT-PCR preparation of mastermix | practical | 0.5 |
3 | Real-time RT-PCR addition of extracted RNA template | practical | 0.5 |
3 | Real-time RT-PCR setup of assay (Cepheid Smartcycler II) | practical | 1 |
3 | Real-time RT-PCR comprehension and analysis of data | practical | 1 |
3 | Written assessment on SOPs | written assessment | 1 |
4 | ‘Scenario day’: independent running of laboratory: processing of batches of mock samples spiked with EBOV ampliconsa | practical & psychological | 4 |
4 | Packaging of samples for shipping as UN2814 | demonstration | 0.5 |
4 | Utilizing the Cepheid GeneXpert | demonstration | 1 |
4 | Spills in FFI and safe clean up | practical | 1 |
4 | Measurement of blood chemistry—Fuji Dri-Chem NX 500 | demonstration | 0.5 |
4 | Measurement of blood clotting—Hemocron Signature Elite | demonstration | 0.5 |
4 | Measurement of WB count—Horiba ABX Micro ES 60 | demonstration | 0.5 |
5 | Analysis of RT-PCR results from both groups ‘Scenario day’ | group discussion | 1 |
5 | Analysis of ‘Scenario day’ and wash-up of course | group discussion | 2.5 |
5 | Skype call with in country team leader | group discussion | 0.5 |
5 | Completion of training records and certificates | training record | 0.5 |
5 | Additional laboratory practice session by requestb | practical | 2 |
If a volunteer was deemed not to have the necessary team worker and laboratory skills or competence in safety critical essential SOPS, they were not deployed. Trainers discussed the performance of each volunteer and gave feedback to the team leaders to help them support and successfully deploy their team. The training programme was routinely evaluated to improve the training by making adaptations to delivery or content to meet the needs of the volunteers and to mirror the reality of events on the ground in Sierra Leone.
A maximum of 16 volunteers were trained per course, and were subdivided into smaller groups of 2 or 4 for hands-on SOP training. Training included specific theoretical, practical and psychological sessions.
Day one comprised familiarization with their teams and deployment arrangements including vaccination updates, visa processing and prerequisite duties required from the respective NGOs. Lectures on Ebola with the latest updates and on the RAs covering their deployment followed. As the epidemic progressed, an eLearning module covering the training material was developed by NADP training and emailed to volunteers for pre-familiarisation with the content prior to attending the training at PHE–Porton. An afternoon session introduced PPE and a practical session on donning and doffing PPE correctly. Prior to receiving an informal discussion on the psychological effects of working in an outbreak environment from individuals with previous experience of such events, volunteers were familiarized with the laboratory and COP. As the outbreak continued, volunteers returning from the PHE laboratories in Sierra Leone shared their experiences with those preparing to deploy.
Day two introduced the concept of good microbiological practice. Volunteers performed a basic laboratory task on the bench to ensure they could pipette accurately without contamination in a precise manner. The same task was then carried out within a FFI to ensure the volunteer could maintain accuracy and aseptic technique when their dexterity was challenged and could appreciate the additional time required when operating in an FFI. Volunteers were fully familiarized with the components and pre-use checks performed on an FFI and carried out troubleshooting techniques including safe glove change and isolator sleeve-repair and replacement. On the afternoon of day 2, groups were further divided into groups of 4 for specific SOP training. Group 1 were trained in sample reception, group 2 in processing samples in the FFI, group 3 in nucleic acid extraction and group 4 in PCR and analysis of results. Each group was trained in rotation on two of these sessions.
Day three training continued the practical SOP training from day 2 with each group of four continuing their rotation of the remaining two SOP sessions. On the afternoon of day 3, volunteers were trained on three blood analysis platforms: Fuji Dri-Chem NX 500, Horiba ABX Micro ES 60 and Hemocron Signature Elite in addition to any novel diagnostic equipment that was undergoing evaluation such as the Cepheid GeneXpert (table 2). Volunteers then completed a written assessment on the SOPs they had been trained in.
Day four training ‘Scenario day’ involved volunteers being divided into two teams of eight and running the laboratory independently in two shifts as if they were in Sierra Leone. Training staff provided a range of mock samples, selectively spiked with known concentrations of EBOV control DNA. It was the role of the teams to process these samples to completion, providing results to the ETC staff (NADP trainers) within an acceptable timeframe. Throughout day four volunteers were challenged to a combination of simulated events; these included power cuts, working in uncomfortably high temperatures, spills of infectious material outside of the FFI, uncontrolled access to the laboratory by a disoriented suspect patient, unannounced visits from government dignitaries and workmen, unapproved visits from the media, excessive pressure from health officials for results, disconcerting sound effects, mis-labelled patient samples, leaking patient samples, opaque containers containing mock sharps, contaminated paperwork, false-negative samples, consumable shortage and excessive staff and understaffing. Each team carried out either a morning or afternoon shift. When not on shift, the team were provided with hands-on training in a separate training area on dealing with spills within an FFI, and the processes involved in safely cleaning up and disposing of infectious material. Additionally, a session was provided on the safe, appropriate storage of samples and instruction in correct packaging for possible transport.
Day five focused on interpretation of qRT-PCR results generated during day 4, database analysis and an extensive reflective and troubleshooting session to discuss the week's events. This also included guidance on appropriate behaviour expected from volunteers when abroad representing their institution and guidance on the acceptable use of social media and photography. The afternoon was left open for volunteers to return to the laboratory to practice further on any of the processes they were not confident in running independently. When possible a conference call was held with team leaders currently deployed so volunteers could get first hand, up-to-date information. The different training modules completed by the volunteers can be seen in table 2. Volunteers then completed their training records and course feedback and received certificates of completion.
14. Outcomes of the field laboratories
The total number of samples tested by all field and mobile laboratories deployed during the outbreak including the timeline of when the three PHE laboratories started testing samples can be seen in figure 6. The Kerry Town laboratory tested 6148 samples during the 12 months the ETC remained operational. Of these, 1209 (19.7%) were EBOV positive and 4738 (77%) EBOV negative, 201 (3.3%) were indeterminate or required repeated testing due to poor sample condition (figure 7a). Port Loko tested 18 458 samples of which 936 (5.1%) were EBOV positive and 16 626 (90.1%) EBOV negative, 896 (4.8%) were indeterminate or requiring repeated testing due to poor sample condition (figure 7b). Makeni processed 26 117 samples to date, inclusive of the samples it has tested as a Legacy laboratory. These include 325 (1.2%) EBOV positive, 25 045 (95.9%) EBOV negative, and 747 (2.9%) indeterminate or requiring repeated testing due to poor sample condition (figure 7c). A mobile PHE laboratory established in October 2015 and still operating from Kenema General hospital at the time of publication tested 4800 samples, of which none have tested positive for EBOV, 4745 (98.9%) tested EBOV negative and 55 (1.1%) indeterminate or requiring repeated testing due to poor sample condition (figure 7d). Sample types tested in each PHE field laboratory correlated to the progression of the epidemic. As the epidemic began to wane, the number of WB samples tested reduced proportionally with the number of ETC inpatients. Contrarily, the number of community samples, made up predominantly of buccal swabs, increased proportionally to the intensification of contact tracing (figure 8).
Figure 6. Number of confirmed EVD cases, total EVD cases and samples tested depicted by epidemiological week. The green arrows represent the date of first testing at each of the three PHE laboratories. (Adapted from [16].) Figure 7. (a) Samples tested by each of the PHE Field laboratories, Kerry Town (2014–2015). (b) Samples tested for EBOV at PHE EVD Laboratory, Port Loko (2014–2015). (c) Samples tested for EBOV at PHE EVD Laboratory, Makeni (2014–2016). (d) Samples tested for EBOV at PHE EVD mobile laboratory, Kenema (2015–2016). Figure 8. Sample types tested in each of the PHE field laboratories, Sierra Leone. October 2014 to July 2016. *PHE Legacy Lab, †PHE mobile laboratory.
15. Discussion and lessons learned for future outbreaks
The principal objective of the pre-deployment training was to prepare volunteer scientists to work in the newly built PHE managed laboratories supporting the ETCs in Sierra Leone. A comprehensive training programme was developed and implemented by the NADP Training team within a very short timeframe. This was possible due to the presence of a team of professionals highly experienced in training scientists from the UK and internationally on practical applications of biosafety and pathogen detection. The NADP training team were also advantaged as a member of the team had previously received training from the EM Lab and led diagnostic EM Lab teams in Guinea prior to the UK response to the West African EVD epidemic. As the epidemic developed, other members of the NADP training team also deployed to run the field laboratories and to train local scientists in Sierra Leone to work in the Legacy laboratories.
The training programme combined the use of some tried and tested concepts from the EM Lab training with several practical aspects designed to accommodate the volunteers lacking in knowledge of molecular diagnostics and the clinical governance associated with working with patient samples. Additionally, volunteers required distinct and specific training in troubleshooting and carrying out basic repairs on the equipment. A powerful component of the training programme was the ‘scenario day’ when volunteers had the opportunity to run the laboratory as if they were in one of the Sierra Leone laboratories. Drawing on challenging situations encountered by previous deployments to the EM Lab and by NADP trainers deployed early in the PHE response, these were incorporated into the training allowing for a highly realistic worst case scenario working environment, and enabling volunteers to deal with these in a safe simulated environment.
Feedback was obtained from volunteers and those deployed to allow the course to be continually updated to reflect the current situation and learn from events in the field. Many found the amount of new information given in the course was high and the course intense. An e-learning module to complement the course was designed and made available to assist in familiarisation of the SOPs and RAs prior to volunteers attending the training. The e-learning was also offered to volunteers who redeployed so they could familiarize themselves with any updates and new equipment being used prior to their redeployment.
During the training the trainers evaluated volunteer scientists and provided guidance on the attributes that would help them to have a successful deployment. The commitment to deploy from microbiologists within PHE was high, for example 50 staff from the research department deployed to PHE or EM Labs during the outbreak, many deploying on multiple occasions. PHE alone could not provide enough staff while continuing to provide its service to the UK, and high numbers of volunteers from NHS, government laboratories and private companies also deployed. In total 402 volunteers deployed to Sierra Leone during the 2013–2016 outbreak, several on multiple redeployments.
In addition to providing a diagnostic service, both the PHE Field laboratories and PHE involvement with the EM Lab supported many studies and clinical trials aimed at improving the diagnostic capability and understanding the immunological markers, virus evolution, virulence determinants and transmission of EBOV [9–11,17–20]. Looking ahead, with DFID funding under the Resilient Zero programme it is planned that all three initial PHE laboratories will transition into Legacy Laboratories (at Makeni, Bo and the Connaught Hospital in Freetown; table 3 and figure 1) and provide a capability to train local staff in testing for a wide range of pathogens of public health concern, in support of the Sierra Leonean recovery plan.
laboratory location | date opened | date closed | no. samples tested |
---|---|---|---|
Kerry Town | October 2014 | November 2015 | 5947 |
Port Loko | December 2014 | December 2015 | 18 458 |
Makenib | December 2014 | legacy | 26 117c |
Kenemaa | October 2015 | August 2016 | 4800 |
Bob | August 2016 | legacy | n.a. |
Connaughtb | pending | legacy | n.a. |
total | 52 034 |
When addressing the lessons learned from this epidemic it is clear that the scale of the epidemic could have been dramatically reduced if initial response efforts had been further supported and the international response had been initiated at an earlier time point. To prepare for future epidemics, the UK government announced at the G7 German summit in June 2015 that Britain would establish a rapid support team deployable to help control future epidemics of public health concern. The wealth of talented scientists prepared to deploy to epidemic situations was evidenced in the recruitment of the volunteers trained during this epidemic. For Britain to take a full role in line with European and international counterparts in response to an infectious disease emergency, an interdisciplinary team that includes trained microbiologists with the laboratory capacity to provide a rapid validated diagnostic service at short notice is essential.
Authors' contributions
C.H.L., J.A.S., S.M.L., M.W.C., A.L., S.F., C.S., S.S., N.J.S. and J.B. carried out conception and design. C.H.L., J.A.S., D.B., A.L., S.F., A.S., N.J.S. and L.E. are responsible for acquisition of data. C.H.L., A.S., T.B. and J.A.S. analysed and interpreted of data. C.H.L., S.H., J.A.S. and A.S. helped in drafting the article and C.H.L., A.S., J.A.S., M.W.C., T.B. and D.B. helped in review.
Competing interests
We declare we have no competing interests.
Funding
This research was funded by DFID.
Acknowledgements
Andy Simpson, Jenny Warner and Neill Keppie for laboratory data retrieval. Rory Miles, PHE Ebola deployment team, EM Lab consortium, Kilian Stoeker, Roman Wolfel, PHE Training volunteers and return deployees, all of the volunteers who attended the Training course.