Sentinel surveillance for influenza A viruses in Lahore District Pakistan in flu season 2015–2016

Lahore is the capital city of the Punjab province of Pakistan (31°32′59″N 74°20′37″E) with a population of 11,119,985 [19]. The district is bounded on the north and west by the Sheikhupura District, on the south by the Kasur District and on the east by the Indian border (Fig. 1). Lahore has a semi-arid climate and experiences four distinct seasons i.e. summer, winter, autumn and spring [20] with June being the hottest month and January the coolest characterized by intense fog and smog, especially over the past 2–3 years [21].

The Lahore General Hospital (LGH) is a major public sector hospital and referral center for the population of Lahore as well as neighboring districts and includes emergency and outpatient clinics and is affiliated with Post Graduate Medical Institute (PGMI) Lahore. The capacity of LGH is 1600 beds. Approximately 14,00,000 outpatients visit various outdoor clinics of the hospital e.g. Medical unit I, II, III, Pediatrics, Neurology, Hepatic, Urology, Orthopedic, ENT, Gynecology etc. annually while the number of outpatient visits is 4000–4500 daily. The number of patients visiting emergency department is 2500–3000 daily. The hospital has 14 ventilators, 10 ICU beds, and 20 isolation beds (https://​lgh.​punjab.​gov.​pk/​clinical_​services).

We prospectively enrolled outpatients with influenza like illness (ILI) and inpatients with severe acute respiratory infection (SARI) according to WHO definition (see below) from those visiting a designated seasonal influenza desk for symptom screening (located within the LGH outpatient department) from Monday to Saturday during official working hours from October 2015 to May 2016. Due to logistic and access issues, we selected one sentinel hospital site from 14 Tertiary Care hospitals and sampled 5–6 patients visiting the hospital on each day, who fulfilled the ILI and/or SARI definitions. These 5–6 patients were selected conveniently from those patients, who visited the hospital in the early hours. Eligible patients were provided with an explanation of the study rationale and sample collection procedures and invited to participate voluntarily. The Independent Ethics Committee (IEC) Bioequivalence Study (Be St) Center, University of Veterinary and Animal Sciences, Lahore (Letter no. IEC-438) approved study protocol.

The standardized and January 2014-revised WHO ILI and SARI case definitions were used. A case of ILI was defined as an outpatient with history of fever (measured temperature ≥ 38 °C) during the past 10 days, and cough without the need of hospitalization. The SARI cases were defined as patients with an acute respiratory illness with onset during the previous 10 days requiring at least an overnight hospitalization and having history of fever or measured temperature of ≥ 38 °C, and cough [22]. Those who did not fulfill the case definition criteria or refused to participate were excluded from the study.

The geographical locations (street addresses) of enrolled patients were located on Google maps. A dot map of distribution of positive and negative cases of influenza A was produced using QGIS software version 3.2 (Open Source Geospatial Foundation Project, Boston, MA, USA, available at http://​qgis.​osgeo.​org).

Using a pretested questionnaire (Additional file 1), epidemiological data about socio-demographics, co-morbidities, exposure status, vaccination, and travel history was gathered from enrolled patients in a face-to-face interview in the local language. A unique coded identification number was given to each questionnaire and specimen tube from the same patients to keep the data confidential.

A registered clinical officer collected nasopharyngeal or oropharyngeal swabs from willing participants [23]. The specimen was immediately transferred into a cryovial tube containing 2-3 ml of viral transport medium. All collected specimens were transported at 4 °C to the Disease Surveillance Laboratory, Department of Epidemiology & Public Health, University of Veterinary and Animal Sciences, Lahore, Pakistan. On arrival in the laboratory, all specimens were divided into small aliquots in a biosafety level-II cabinet (BSL-2) and were stored at or below − 70 °C before shipment to the WHO Collaborating Center for Studies on the Ecology of Influenza, St. Jude Children's Research Hospital, Memphis, Tennessee, USA for influenza testing [23, 24] by using real-time RT-PCR (rRT-PCR) protocol as described by the US Centers for Disease Control and Prevention [25] for IAV. Positive IAV specimens were further subtyped for seasonal H3N2 and H1N1pdm09 by rRT-PCR using specific primers, probes [26].

Questionnaire data were entered into Epidata version 3.1 (available at http://​www.​epidata.​dk) and validated for errors and inconsistencies. We used R software version 3.2.3 [27] for analysis of socio-demographic and clinical characteristics. Graphs were generated to present spatio-temporal trends of IAV and subtypes (seasonal H3N2 and H1N1pdm09). The denominator for calculation of various proportions was the total number of cases (all diseased) who visited the hospital during study period (n = 513,126) and total number of ILI and SARI patient enrolled in study (n = 311).

From October 2015 to May 2016, a total of 513,126 patients reported to the LGH surveillance desk designated for infectious diseases. A total of 4568 patients visited the designated seasonal influenza desk with ILI or SARI symptoms. Of the 4568 patients, 311 were invited to participate and were enrolled in study. Among 311, 284 were ILI (91.3%) and 27 were SARI (8.7%) patients. The mean age among all enrolled patients was 29.20 years (median 27 years; range: 0.33–80 years). Most ILI (70.1%, mean = 29 years) and SARI patients (66.67%, mean = 29 years) were under the age of 30 years. The majority of cases were females (n = 202, 65%). Among these 202, 184 were ILI patients and 18 were SARI cases. Out of these 202 females, 140 were married and among these, 91.4% (128/140) were ILI patients and 8.6% (12/140) were SARI patients. Among married ILI female patients (n = 128), 3 (2.3%) were in first trimester of pregnancy, and 4 (3.1%) were in second trimester of pregnancy; none were in the third trimester. About 94.5% (n = 121) female ILI patients were married but not pregnant. Among married SARI female patients (n = 12), 1 (8.3%) was in the second trimester and 1 (8.3%) was in the third trimester. The number of ILI and SARI patients was higher among housewife’s compared to other occupations. In overall, a single-family system was more common (187/311: 60.1%) and families had an average monthly income of US$172. No deaths were recorded among enrolled cases. About 31.5% of patients (98/311) reported that they self-medicated with antibiotics (available over the counter in Pakistan). Among ILI patients, 27% reported self-medication with antibiotics and among SARI, 29.6% reported use of antibiotics. No patient reported the use of any antiviral drug. Out of 311 patients, 99.4% (n = 304) were unaware of the availability of any influenza vaccine. Only two patients (0.6%) knew about the seasonal influenza vaccine and one (0.3%) was vaccinated against influenza and they were among ILI cases. Among ILI patients, 16 (5.6%) were vaccinated with pneumococcal vaccine, while in SARI cases, one (3.7%) was vaccinated with pneumococcal vaccine. Majority of ILI (80.9%) and SARI (88.8%) had no travel history. Similarly, majority of ILI (87.6%) and SARI (88.8%) reported no contact with poultry. Most of the ILI (94.3%) and SARI (92.5%) did not used public transport (Tables 1 and 2).

Among all enrolled individuals, 16% (50/311) were positive for IAV. Of these, 80% (n = 40) were from ILI and 20% (n = 10) were from SARI groups (Fig. 2). Among ILI cases, 14% (40/284) were IAV positive while among SARI cases 27% (10/27) were IAV positive.

The proportion of respiratory illnesses (n = 4568) among all patients visiting the sentinel hospital during study period (n = 5,13,126) was 0.9% (95% Confidence Interval [CI]: 0.86–0.92). The proportion of enrolled ILI (284/311) was 91% (95% CI: 87.62–94.20) and enrolled SARI (27/311) was 8.68% (95% CI: 5.79–12.38). The maximum monthly number of ILI (20%) and SARI (26%) patients was seen in February (Fig. 3).

The proportion of IAV positive among all enrolled was 16% (50/311; 95% CI: 12.17–20.64). Among the 50 IAV positive cases, virus subtyping detected 14 (28%) as H1N1pdm09 and 15 (30%) as seasonal H3N2; the remaining 21 (42%) were untyped (Fig. 4). The detection of IAV was limited to winter months (December–February). The highest proportion (42%) of positive IAV cases was recorded during December 2015. (Fig. 4).

Among the ILI and SARI cases, 22.5% and 50% were H1N1pdm09 positive, 32.5% and 20% were seasonal H3N2 positive and 45% and 30% were untyped IAV positive respectively.

Among ILI patients, the number of IAV positives was highest among the 16–30 years (55%, 22/40). Among IAV positive ILI cases, 55% were male and 45% were female. No pregnant ILI females (n = 7) were IAV positive (Table 1).

Among SARI cases, the number of IAV positives was highest in 16–30 age group (70%, 7/10). Among IAV positive SARI cases 70% were female and 30% were male (Table 2). Both SARI pregnant females were IAV positive.

Ongoing or history of fever was the most common symptom reported in the ILI (91%) and SARI (96%) groups. Cough (85%) and Sore throat (78.5%) were the other common clinical symptoms reported by ILI cases. Cough (89%), sore throat (67%), running nose (74%), headache (74%) and muscle pain (74%) were frequently reported symptoms by SARI cases (Fig. 5, Table 3).

The most common ILI co-morbidities were hepatitis C (5.2%), tuberculosis (4.2%) and diabetes (4.2%), while the most common co-morbidities in SARI cases were diabetes (7.4%), hepatitis C (7.4%) and obesity (7.4%). In all IAV positive cases, the most common comorbidities were hepatitis C (4%), obesity (4%) and tuberculosis (6%) (Table 3).

The majority of the patients reported to the hospital on day three post symptoms onset (66/311) (interquartile range [IQR] = 2–6 days) with 14 of these IAV positive (14/66).

About 93.2% of the reported cases of ILI and SARI (290/311) lived in the Lahore district (Fig. 6); the remaining 21 were from other districts of Punjab province. Among the cases from Lahore district, 265 were ILI patients and 25 were SARI patients. Among IAV positive cases (n = 50), 47 were reported from Lahore district (94%).