The initial literature search yielded 1641 records: 956 publications in Scopus, 668 in PubMed, 14 in OVID, 2 in Cochrane, 1 in NLM Gateway, and 0 in DARE. All records were compiled into an EndNote X5 (Thomson Reuters, New York, NY) library. The library was then inspected for duplicates, and 1156 records were retained. These records were then checked for relevance, and only records that were research papers on influenza in the EMR were retained regardless of whether the authors were based in the region. After this step, 510 records remained, were compiled in a new EndNote library sorted by country, and were included in the analysis. A spreadsheet was constructed with the following data fields: author, title, year, reference, country, type, population, intervention, comparison, and outcome. Publications were classified according to type: editorial (including opinion pieces); epidemiology; epizootiology; healthcare management; knowledge, attitude, and behavior (KAB); modeling; reporting; review; surveillance; and basic virology research.
By country, type, and date
Figure 1 shows the distribution of publications by country. Four countries appear to have active influenza research groups producing more than 75% of the published research in the region. Iran leads these countries with 135 records (26%), followed by Egypt with 110 records (21%), Saudi Arabia with 92 records (19%), and Pakistan with 46 records (9%). Iraq, Kuwait, Lebanon, Tunisia, UAE, Jordan, and Morocco had between 10 and 15 records each (2–3%). Other EMR countries had fewer than 10 records (<2%), and no records were found for Djibouti, Somalia, or Syria. Only 22 records (4%) were shared between two or more countries.
Most of the published research (29%) was epidemiologic research (Figure 2). On close inspection, most of these were descriptive clinical epidemiology papers describing cases of influenza in hospitalized populations, complications of influenza, or influenza as a cause, among other agents, of respiratory tract infections. Very few had an analytical epidemiology design, were controlled, or were clinical trials.
Basic virology papers constituted a major portion of reviewed research (28%). Most of these papers experimented with H5N1, H9N2, or pandemic H1N1 viruses. Most research was directed toward molecular and phylogenetic analyses of these viruses. Some explored the pathogenicity of avian influenza viruses in poultry, and a few experimented with antiviral resistance, development of vaccines and diagnostic techniques, and vaccine efficacy, all categories of research that have immediate relevance to public and veterinary health.
Epizootiology research accounted for 63 records (12%). Most papers found described the prevalence of H5 or H9 viruses among poultry and wild birds. A handful of papers looked at other species such as horses, dogs, swine, and cats.
Review, editorial, opinion, case reporting, and KAB papers constituted about 27% of published research. Although such papers may have immediate relevance to policy and decision making, they remain descriptive and would be of limited research value. Papers showing results of surveillance among birds or humans, as well as research on the burden of disease, were very limited in number.
Figure 3 shows the distribution of records by date. Papers were classified by year for papers published as of 2006, when H5N1 started spreading in the region. Papers between 2000 and 2005 were grouped together. Papers between 1950 and 1999 were grouped by decade, and papers before 1950 were grouped together. It is worth noting that the first record found in this review dates back to 1894. This was an article published in the British Medical Journal describing cases of influenza in Egypt. Papers from the 1950s, 1960s, and 1970s described the prevalence of various influenza viruses, such as the Asian flu and Hong Kong flu, in human populations. Although very few papers were from the 1980s, these papers were the first to study animal influenza and the potential antiviral activity of some compounds. Influenza research activity, although diverse, remained rare in the 1990s and through 2005.
In 2005, H5N1 viruses arrived in some EMR countries. This boosted influenza research, especially in affected countries. More research was conducted between 2006 and 2008 than during any previous time. The quality of research improved and was more diversified, and virology papers were frequent. Within a few years of avian influenza emergence in the EMR, the pandemic H1N1 of 2009 arose. This also helped boost influenza research in the region, especially because the Hajj was a major concern due to the potential of rapid transmission of the novel virus in a vulnerable group. Influenza research in the region reached its highest level in 2010 (123 papers).
Searching the IRD for published influenza sequences from EMR countries yielded 3099 influenza A sequences, 56 influenza B sequences, and no influenza C sequences. The Epiflu search yielded 2259 influenza A sequences, 125 influenza B sequences, and no influenza C sequences. About 500 influenza A sequences were found on Epiflu but not on IRD, and about 1340 were found on IRD but not on Epiflu. Because IRD provided a more complete dataset, the results of the search of that database are presented here. Figure 4 shows the distribution of influenza A sequences by country. The largest proportion was from Egypt with 1280 published sequences. These sequences included avian, equine, and human viruses, but the vast majority were from avian species due to the endemicity of H5N1 in Egypt. Iran contributed 533 sequences, about half of which were from human influenza A viruses and half from avian viruses, especially H9N2. Pakistan ranked third with 422 sequences, all of which were avian influenza viruses, especially subtypes H7 and H9. Other EMR countries published fewer sequences, and none were found from Libya, Somalia, Syria, or Yemen.
Influenza A sequences were from viruses from as early as 1968, but sequence data significantly increased starting in 2006 with the arrival of H5N1 to the area. Of the 3099 sequences, about 800 were for human viruses, and the rest were from avian hosts. Among avian viruses, H5N1 had 1469 sequences, H9N2 had 522, and H7N3 had 227. Sequences from other subtypes (H6N2, H7N7, H8N4, H10N7, and H11) were less frequently reported. Although most of the sequences were for the HA gene (1350) or the NA gene (629), one-third of the published sequences were from internal genes.
Few influenza B sequences were reported, and they were from Egypt, Iran, Lebanon, Morocco, Oman, and Saudi Arabia.
Reviewing virologic data available on FluNet between week 19 of 2011 and week 19 of 2012 (http://www.who.int/influenza/gisrs_laboratory/flunet/en/) revealed that the following EMR countries were contributing data: Afghanistan, Bahrain, Egypt, Iran, Iraq, Jordan, Morocco, Oman, Pakistan, Qatar, Syria, and Tunisia. Such data are typically contributed by the national influenza centers (NICs) or the central laboratories of reporting countries. Data available indicate that there were 419 confirmed cases of pandemic H1N1, mostly from Qatar, Pakistan, and Oman (237, 106, and 65, respectively). H3N2 viruses were also in circulation, with 1788 confirmed cases reported, most from Iran, Egypt, and Morocco (882, 237, and 236, respectively). The same can be said about influenza B viruses as regional countries reported 371 cases, mostly from Qatar, Oman, and Iran (118, 89, and 79 cases, respectively).
Data in Figure 5 show that the incidence of influenza infection in humans in the EMR followed a seasonal pattern, peaking in the fall and winter weeks. However, influenza cases continued to be reported throughout the period between week 19 of 2011 and week 19 of 2012. During that period, most of the infections were due to H3 viruses, especially toward the end of 2011. H1 viruses started to become more dominant in week 10 of 2012. Influenza B viruses continued to cause cases throughout the year.