Our search combined electronic and manual searching methods to capture published journal articles, conference papers, and grey literature. We consulted four electronic databases: PubMed, Web of Science, Index Medicus for the Eastern Mediterranean (IMEMR), and CINAHL. We treated Google Scholar as part of our manual searching strategy owing to its limitations, namely its 256-characters search limit and maximum display of first 1000 records without specification of ordering [26]. We formulated a search strategy in consultation with a specialized librarian. The strategy had two central concepts: snake bites and EMR countries and incorporated a combination of MeSH terms and text words. We tested, refined, and tailored the strategy to match each included database (provided in S2 Supplement). The search was executed on three databases from February 13 to 14, 2023, and expanded on April 6, 2023, to include the Index Medicus for the Eastern Mediterranean Region (IMEMR).

Manual searching methods included snowballing of search-captured literature reviews; expert consultation to locate further articles and data sources; searching websites of national Ministries of Health to obtain grey literature e.g., the most recent statistical health report containing data on snake bites; and utilizing Google Advanced Search tool with English, Arabic, and French keywords.

Our search was limited to publications from 2000 to 2023 on the premise that older articles may be outdated relative to present data sources.

We exported records to the citation management software (Zotero 6.0) so that duplicates could be removed electronically and checked manually.

We included only primary records involving confirmed snakebite human victims regardless of envenoming status. We therefore excluded cross-sectional studies on knowledge, perspectives, and experiences; in vivo and in vitro studies; modeling studies; and literature reviews. Furthermore, we limited inclusion to articles published in Arabic, English, French, and Persian since they are the major publishing languages of health research in EMR [27]. Inclusion and exclusion criteria are further specified in Box 1 below.

One reviewer first performed screening of titles and abstracts against our inclusion and exclusion criteria. If the abstract was not available, the article was included in the subsequent stage. Relevant records then underwent full-text retrieval and screening. Weekly meetings with a second researcher took place to discuss and resolve emerging uncertainties during this stage.

We contacted 22 authors and journals via email to (i) request articles that could not be retrieved, or to (ii) make enquiries if information was missing e.g., the country of cases was not specified. All non-respondents were sent second follow-up emails, and the final response rate was 50% (11/22).

Data charting took place via a standardized Excel extraction form. The process was largely iterative i.e., some data items were later added [28], and multi-staged. One reviewer first extracted study characteristics from all but Persian studies as another reviewer was sought for those. The reviewers collected data on country focus, language, date of publication, journal/conference name, type of study as per Bolon et al. [29] with adaptations, study aim, first author, and academic affiliations of authors. Second, reviewers collected and organized data numerically and thematically around three a priori pillars. The first pillar was the burden of snake bites and encompassed total number of cases reported, number of cases included in the analysis, deaths, permanent sequelae, disability, sociodemographic profile of cases, and national-level incidence and case fatality. Second pillar was snake species i.e., offending species and snake identification methods. For this, one reviewer cross-referenced offending species from older publications against recent taxonomic revisions to ensure consistency in species identification. Further, we compared the species reported in the literature with WHO’s Snakebite Information and Data Platform [https://snbdatainfo.who.int/] and herpetology and toxicology guides on the Region [30–32]. The third pillar covered types and manufacturers of antivenoms used, specificity (corroborated with WHO’s platform), number of cases receiving antivenom, and related antivenom challenges. A second reviewer was sought in case of uncertainty.

The quality of included studies was not appraised, in following Arksey and O’Malley framework [23]. We used frequencies, percentages, tables, and descriptive narrative summaries to report findings.

Of the included 196 records, there were 182 (93%) articles published in 115 journals, 10 conference abstracts (5%), and four official health reports (2%). Around 72% (141/196) of included records were published in the second two quarters of the specified timeframe (2012–2023). Most publications were in English (166/196, 85%), followed by French (24/196, 12%), Arabic (4/196, 2%) and Persian (2/196, 1%).

More than half (54%) of the publications focused on Iran (37/196, 19%), Morocco (35/196, 18%), and Pakistan (35/196, 18%). This was followed by Saudi Arabia (15/196, 8%), Sudan (13/196, 7%), and Oman (11/196, 6%). While each of the remaining 16 countries had ≤6 articles, we could find only two articles (1%) each for Kuwait, Somalia, and Libya, and no records for Syria or Bahrain. Two studies covered more than one EMR country. Fig 3 shows the number of records per country.

In terms of methodological focus, a third of included records (67/196, 34%) were case reports or case series. Retrospective and prospective facility-based studies constituted 30% (58/196) and 21% (41/196) respectively. Only 9% (18/196) of publications were national registry-based, and about 4% (7/196) were clinical trials and evaluation studies. Study characteristics are summarized in Table 1.

Authors’ affiliations spanned 261 institutes. The Anti-Poison and Pharmacovigilance Center (Morocco), Ibn Tofail University (Morocco) and the Mashhad University of Medical Sciences (Iran) top-ranked with 15, 12, and 9 publications respectively. Around 19% (37/192) of journal and conference papers had at least one author affiliating with an organization outside the region.

There were 67 case reports and case series with varied focus and no discernable pattern. For the remaining records, close to half aimed to investigate the clinical and epidemiological features of snake bites irrespective of envenoming status (57/129, 44%), whereas 17% exclusively addressed snakebite envenoming (22/129). Some records had broader or specialized focus with aims spanning across poisoning (12/129, 9%), renal complications (11/129, 9%), animal bites (6/129, 5%), coagulopathy (3/129, 2%), and economic burden of bites (2/129, 2%). There were seven studies (5%) that aimed to examine antivenoms’ effectiveness, efficacy, safety, and/or cost.

A total of 170,626 snake bites from 20 countries were reported in papers published between 2000 and 2023. Mortality was not measured in 25 publications (13%), but there were 2,551 deaths reported across the remaining records.

Permanent sequelae were present in 15% (29/196) of included publications. There were another 82 publications (42%) which did not mention permanent sequelae, and thus neither reported nor ruled out these outcomes. The main sequalae that were reported were gangrene and amputations (14 studies); deformed extremities and motor weakness (7 studies); chronic renal damage (3 studies); and visual impairment (3 studies). Other less frequent sequalae were erectile dysfunction and permanent bladder areflexia [34], and chronic post-traumatic stress disorder [35] reported in one record each. Furthermore, three studies collected a disability variable [36–38] with only one study by Akbar et al. providing a working definition of disability as: “development of permanent organ dysfunction (hemiplegia due to intracerebral haemorrhage or permanent loss of higher mental functions due to hypoxic brain damage resulting from respiratory insufficiency)” [36].

Regarding the sociodemographic profiles of snakebite victims, there was a higher frequency in males noted in 97/100 of the population-level records (excluding case reports and case series, and those that did not report on sex). Publications used different ways to report age distribution of victims i.e., mean, median, and frequencies, as well as different age grouping which makes it difficult to summarize. For example, three studies grouped victims as children and adults older than 15 years [39–41]. Snake bites mostly affected rural populace in 39 of the 40 records indicating rural/urban residency. Six studies from Saudi Arabia, Kuwait, and Oman collected data on the nationality of patients i.e., nationals versus non-nationals, with mixed findings from Saudi Arabia [41–43] and Oman [44,45] and higher frequency among non-nationals in one Kuwaiti study [46]. Three prospective studies contained data on education status, showing higher occurrence of snake bites among the least educated subpopulations [47–49]. One study measured socioeconomic status of victims and another reported on family income with findings of higher frequency among those living in poverty or with lower income [38,47].

Six countries in the region collected national-level data on snakebite burden. These were Iran, Saudi Arabia, Palestine, Lebanon, Sudan, and Morocco. The latter four countries covered snake bites in their periodic statistical reports [50–53]. Among the 18 national-level records included in the review, seven calculated an incidence risk. It was lowest in Morocco at 0.34/100,000 for the years 1980–2008 [54]. It was highest in Sudan at 47/100,000 in 2014 [55]. The highest country-wise case fatality rate was observed in Morocco at 5.66% for the years 1992–2007 [56]; however, it decreased to 1.62% in 2021 [53].

The offending snake was not identified in 109 of the 196 included records (56%). The remaining 87 publications proposed identifications for some or all the culprit snakes at family and genus levels (30/87, 34%) or at the species level (57/87, 66%). About 37% of these studies (32/87) did not report the method by which the snake was identified. Those that did specify (55/87, 63%) used different methods to do so, with 13 studies using more than a single method. The most common identification method was body examination of killed or captured snakes in 19 records. Second was victims or witnesses describing or self-identifying the snake in 18 records. Alternatively, snake identification was based on clinical presentation and laboratory findings in 12 articles. Photo catalogues and posters were presented to victims to identify the implicated snakes in seven studies, whereas photography of killed or captured snake were brought by patients in six studies. The bites involved a snake collector, a herpetologist, or a snake charmer in six publications.

Identified indigenous snakes belonged to 30 species. Cerastes cerastes were the culprit species in 16 studies and second was Echis carinatus in 10 studies. This was followed by Bitis arietans, Echis coloratus, and Macrovipera lebetina in seven records each. Daboia mauritanica was cited in six records all from Morocco. Also, Naja haje and Daboia palaestinae were reported in five and four publications respectively.

Fifteen species were only reported in one record each and almost all were responsible for 1–5 bites. The only exception was Atractaspis microlepidota being implicated in 23 bites according to one study on 63 snakebite-envenomed children from Sudan [57]. There were one sea snake bite (Hydrophis platurus) inflicted upon a fisherman in Pakistan [58] and one water snake bite (Natrix maura) from Morocco [59]. S4 Supplement provides a table of offending snake species per country as per included records.

Species reported in the literature were consistent with WHO’s Snakebite Information and Data Platform and herpetology and toxicology guides except for three snakes. These were Echis carinatus from Sudan, Vipera berus in Saudi Arabia, and Cerastes vipera from United Arab Emirates mentioned in one record each [57,60,61].

Some or all the victims received antivenoms in 75% of included records (147/196), none was administered in 6% (11/196), and it was not clear in the remaining 19% (38/196). Only 64% of records specified the exact number of cases receiving antivenoms (125/196). Half the publications where antivenoms were administered did not provide information on the type (75/147, 51%). For instance, 27 publications only mentioned that the antivenom was ‘polyvalent’ without providing details on the manufacturer, trade name, and specificity. This lack of information applies to all the records obtained from Sudan, Palestine, Somalia, and Libya.

We counted more than 12 types of antivenoms used in EMR countries. Six types were produced by antivenom production centers from six countries in the Region. These were Pakistan, Iran, Saudi Arabia, Syria, Egypt, and Tunisia. The other types originated from France (FAV-Afrique, Favirept, and Bitis-Echis-Naja), Spain (Inoserp MENA), and India (Serum Institute of India polyvalent and VINS Biosnake).

In Iran, the only types of antivenoms documented in the country’s literature was produced by the Razi Serum and Vaccine Research Institute [62–67]. This Razi polyvalent antivenin was further used in US-military hospitals in Afghanistan [68,69]. For Saudi Arabia, antivenoms were manufactured by Antivenom and Vaccine Production Center of National Guard health affairs [70–72]. Use of the Saudi antivenoms has also been documented in Oman [44,73–75], Qatar [76,77], United Arab Emirates [75], Kuwait [46], Yemen [78], and US-military hospitals in Iraq [79]. Further, the Syrian ‘Antivenom-2’ manufactured by Scientific Studies and Research Center was used in Lebanon [80,81], with one study questioning its ‘efficacy’ in Lebanon as it showed large variation in hospital stays and 67% complications rate requiring intensive care admissions [81].

As for the Tunisian and Egyptian antivenoms, we did not retrieve any records of use outside their countries of origin. In Pakistan, production of the polyvalent produced by the National Institute of Health in Islamabad was insufficient to meet the country’s demand, and Indian antivenoms were imported to close the gap [47,82]. The Pakistani antivenom was claimed to be 2.5 times more effective and less costly than its Indian counterpart as per one trial study [82]. Table 2 summarizes data on antivenoms, including types, manufacturers, specificity, countries of use, and number of citing records.

About 27 publications described challenges relating to antivenoms. Shortages in supply were noted in 12 publications, with two studies from Pakistan and Djibouti highlighting that antivenoms were consequently administered below required doses [83,84]. In contrast, antivenoms were over-administered to patients even in absence of “clinical manifestations of envenomation” according to three studies on Yemen [78], and the United Arab Emirates and Oman [75,85]. Furthermore, the antivenoms used were sometimes not specific to the offending species as per nine studies from seven countries. For example, Abu Baker et al. documented the ‘inappropriate’ use of VINS Biosnake, the only antivenom available in Jordan, against Atractaspis engaddensis and Echis coloratus concluding that: “Lives could have been saved if specific antivenom had been available” [86].