Evaluation of the influenza sentinel surveillance system in the Democratic Republic of Congo, 2012–2015

An influenza sentinel surveillance system (ISSS) was established in DRC in 2006 following the emergence of the highly pathogenic avian influenza A(H5N1) strain in Asia with a high case-fatality rate [11]. The objectives of the DRC-ISSS are to: (i) detect and respond to influenza outbreaks; (ii) assess the proportion of patients meeting the ILI and SARI case definition that is attributable to influenza virus infection; (iii) assess the burden of influenza-associated illness; (iv) monitor the temporal trends of influenza virus circulation; (v) monitor the circulating influenza virus types and subtypes annually; (vi) maintain laboratory capacity for seasonal and avian influenza viruses detection; and (vii) obtain and share clinical samples for annual selection of influenza virus strains for influenza vaccine formulation under the WHO-GISRN. In addition, the data generated through the surveillance system were considered key to potentially inform and promote prevention interventions. The ISSS was coordinated and implemented by three institutions, namely: the Kinshasa School of Public Health (KSPH), the Direction de la Lutte contre les Maladies (DLM) and the Institut National de Recherche Biomédicale (INRB). International stakeholders included the CDC and WHO Country Offices.

Surveillance was designed to be implemented in 6 of 11 provinces of the country. During the review period (2012–2015) the ISSS included 5 of 6 target provinces. Although health facilities were available in all provinces, provinces to be covered were selected based on available financial resources and pre-established criteria. Priority was given to provinces situated at major entry points of the country with significant population density and movement and to those located along the corridors of migratory wild birds. The selected provinces were situated throughout the national territory, with the exception of the northwestern part of the country.

In total 11 sentinel sites (2 clinics implementing ILI surveillance only and 9 hospitals implementing both ILI and SARI surveillance) located in 5 provinces were included in this evaluation (Table 1 and Fig. 1). In the 9 hospitals, ILI surveillance was conducted in the outpatient department; whereas SARI surveillance was conducted in the medical pediatric and adult wards. At each sentinel site trained staff (i.e., doctors, nurses or laboratory technicians) would: (i) screen, identify and enroll patients, (ii) compile individual-level screening and enrollment logbooks for patients meeting the SARI and ILI case definitions, (iii) collect upper respiratory tract (URT) samples from enrolled ILI and SARI cases, (vi) complete the individual case investigation form (CIF) for enrolled ILI and SARI cases, (v) store, package and ship URT samples, (vi) collect weekly and monthly aggregated data on the total number of any, respiratory (including those that met the ILI and SARI case definitions), gastrointestinal and malaria outpatient consultations and hospitalizations; and (vii) liaise with the national influenza surveillance focal points on all matters related to influenza surveillance implemented at the sentinel sites, including requests for surveillance material. No monetary incentive was provided to the influenza focal points. Airtime was provided for communication between sentinel sites and national focal points. Nonetheless, the ISSS in DRC was largely dependent (≥90%) on external funds, especially for sample transportation and testing as well as for supervision activities.

A case of ILI was defined as an ambulatory patient of any age presenting with a recorded temperature ≥ 38 °C and cough or sore throat of duration of ≤7 days. A case of SARI was defined as a hospitalized person who had illness onset within 7 days of admission and who met age-specific clinical inclusion criteria. A case in children aged 2 days to < 5 years included any hospitalized patient with cough or difficulty breathing and at least one of the following danger signs: unable to drink or breastfeed, lethargic, vomits everything, convulsion, chest indrawing or stridor in a calm child. A case in persons aged ≥5 years included any hospitalized patient with fever (≥38 °C), cough and shortness of breath or difficulty breathing. Identification of ILI cases was implemented at the triage area of the participating sites by a combination of clinical examination for relevant symptoms and patient/caregiver interviews. Identification of SARI cases was implemented at the participating wards by bed-to-bed screening of admitted patients and a combination of clinical examination for relevant symptoms, review of medical records and patient/caregiver interviews.

Aggregated weekly data on the total number of identified SARI and ILI cases and those enrolled were sent weekly by short message service (SMS) to the national influenza focal point; whereas the source data collection forms were sent monthly. For enrolled patients surveillance staff completed case report forms that included demographic, clinical and epidemiological information as well as specimens (nasopharyngeal and oropharyngeal swabs). Specimens were placed in the same vial containing universal transport medium, stored at 4–8 °C and transported to the national influenza laboratory (INRB, Kinshasa, DRC) within 72 h of collection for testing. Specimens were tested for influenza A and B viruses using a real-time reverse transcription polymerase chain reaction assay [6]. Influenza A-positive samples were further subtyped [12]. Non-systematic testing for other respiratory viral pathogens including respiratory syncytial virus, human metapneumovirus, parainfluenza virus types 1–4, rhinovirus, coronavirus OC43, 229E, NL63 and HKUI, and adenovirus was also implemented. Verbal informed consent was obtained from all patients prior to data and specimen collection. For children aged < 15 years, verbal consent was obtained from a parent or legal guardian.

The appointed surveillance officers at the sentinel sites were supervised by DLM staff in order to ensure adequate performance in the detection of cases, data collection and collection and storage of samples. During the supervision visit, knowledge, skills and practices of the surveillance officers related to influenza surveillance procedures were evaluated using a standard evaluation checklist. Ad-hoc trainings were conducted to address any deficiency. These supervisions were carried out monthly at the sentinel sites situated in Kinshasa Province and, due to financial constraints, twice per year at the sites situated in the other provinces. Annual refresher trainings were conducted for all personnel involved in ISS at the participating sites. Individual-level laboratory results were communicated monthly to the sentinel sites and weekly, monthly and quarterly reports were generated and shared among relevant stakeholders. However, no thresholds to assess the intensity and impact of the influenza season were established.

We used the CDC guidelines [3, 4] to evaluate the performance of the ISSS in DRC during January 2012 through December 2015. In accordance with the CDC guidelines, the performance of the system was evaluated using eight surveillance attributes: (i) data quality and completeness for key variables, (ii) timeliness, (iii) representativeness, (iv) flexibility, (v) simplicity, (vi) acceptability, (vii) stability and (viii) utility. For each attribute, specific indicators were developed and described using quantitative and/or qualitative methods (Tables 3, 4 and 5). Surveillance databases from syndromic and virological surveillance as well as the laboratory receiving log were analyzed to assess data quality and completeness, timeliness and stability. Furthermore, in order to assess simplicity, acceptability, stability and utility, standardized, pretested and anonymous questionnaires were self-administered by surveillance staff at sentinel sites as well as key informants (i.e., staff involved in ISS or leadership) from the DLM and INRB. All personnel involved in ISS was requested to participate to the questionnaire survey. Three different questionnaires were developed, one for each of the three target groups (i.e. surveillance staff at sentinel sites and staff at the INRB and DLM). Data collected from the surveillance system were also compared with WHO minimum data collection standards for ILI and SARI surveillance [1].

For consistency and comparability of findings we used the evaluation method and scoring system utilized for influenza surveillance evaluations conducted in other African countries [7‐10]. A scale from 1 to 3 was used to provide a score for each quantitative indicator as follows: < 60% scored 1 (weak performance); 60–79% scored 2 (moderate performance); ≥80% scored 3 (good performance) [8]. For qualitative indicators a score was assigned based on the same scale using expert consensus. Thereafter the scores assigned to each indicator were averaged for all indicators evaluated within each attribute to provide an overall score for each surveillance attribute assessed in this study. An overall score for the surveillance system was obtained by averaging the eight mean attribute scores. This evaluation was implemented by personnel from the KSPH (internal evaluators) and CDC (external evaluators) and was not linked to the Joined External Evaluation conducted in DRC in 2018.

Categorical variables were expressed as percentage of outcomes of interest over total observations for each quantitative indicator. Furthermore, 95% confidence intervals for proportions were calculated using the binomial distribution. The statistical analysis was implemented using Stata version 14.2 (StataCorp, College Station, Texas, USA).

During 2012–2015, 16,152 patients with respiratory illness were reported from the 11 sentinel surveillance sites. Of these, 11,737 (72.7%) were outpatient consultations and 4415 (23.3%) were hospital admissions of which 4812 (40.9%) and 2869 (64.9%) met the ILI and SARI case definitions, respectively. Of the 7690 patients eligible for enrollment, 7090 (92.2%) were enrolled in the surveillance system; 4339/4821 (90.0%) and 2751/2869 (95.9%) among patients with ILI and SARI, respectively. Influenza viruses were detected in 597/7090 (8.4%; 95% CI: 7.8–9.1%) samples tested. Of these, 111 (18.6%) were influenza A(H1N1)pdm09 viruses, 209 (35.0%) were influenza A(H3N2) viruses, 27 (4.5%) were influenza A viruses not subtyped and 251 (42.0%) were influenza B viruses (Fig. 1). Influenza viruses were detected in 446/4339 (10.3%; 95% CI: 9.4–11.2%) samples from patients with ILI and in 151/2751 (5.5%; 4.7–6.4%) samples from patients with SARI. Influenza viruses were detected mainly during December to May (Fig. 2). The questionnaire survey was completed by 35/39 (89.7%), 3/4 (75.0%) and 6/6 (100.0%) personnel involved in influenza surveillance at sentinel sites, INRB and DLM, respectively.

The overall mean score for the ISSS in DRC was 2.5 (moderate to good performance out of a 1–3 scale) (Table 2).