A total of 11,587 articles were identified; 3384 from Embase, 3221 from Medline/Ovid, 2393 from Scopus, 1207 from Web of Science, 864 from PubMed, and 518 from Open Grey. None were identified from the updated search. After removal of duplicates, 9829 articles remained. Out of this total, 9758 articles were excluded based on titles and abstracts. Two articles were then added based on a manual search of Google Scholar and reference lists of identified studies. The full-text screening of the articles excluded a further 66 articles for reasons reported in Fig. 1. A total of 8 eligible studies from 7 publications were included in the analysis (see Fig. 1). In the overall quality assessment, four studies scored high, three scored medium, and one scored low. Out of 138 LMIC, which incorporate 37 FCAS, five countries were included in our analysis.
Study characteristics
Three studies were conducted in Lebanon (Beirut and Byblos) [22, 23], two in Iraq (Baghdad and Kurdistan) [24, 25], one in Jordan (Ar Ramtha) [26], one in Palestine (Gaza) [27] and one in Yemen (Sana’a) [28]. The study population (total of 5833 participants) was composed of: 1- Syrian refugees with acute conflict injuries [26], 2- military and civilian populations with conflict-inflicted injuries [24], and 3- either civilians or undetermined populations in the five remaining studies. The results consisted of three retrospective chart reviews [23,24,25], one cross-sectional study [28], one case-control study [27], one nested case-control [22] and two prospective cohorts [22, 26]. All studies were hospital-based but only three were multicenter, and half of the studies were conducted in intensive care units with or without including other hospital wards [22, 24, 27]. Five studies investigated the burden of only one bacterial species [22, 23, 27, 28]. The most studied organism was Acinetobacter baumanii (n = 5), followed by Staphylococcus aureus (n = 4), and Escherichia coli (n = 3). Six studies reported the prevalence of MDR infections which ranged between 37.5 and 88%. However, none of the studies explicitly and comprehensively assessed the COI from resistant infections. Detailed characteristics of the studies included in this review are summarized in Table S2 (see Additional file 1).
Resource consumption
Hospitalization costs
Only one study estimated the average cost of hospitalization, direct cost, due to resistant infections [22]. In this study, the average cost of hospitalization in the ICU due to MDR Acinetobacter infection was estimated to be around $1750 per day. The length of stay (LOS) was reported to increase by up to 2 weeks when patients were diagnosed with device-associated infections caused by the resistant organism, leading to an increase in hospitalization cost by $24,000 for every MDR Acinetobacter infection.
Length of hospitalization
Two studies reported higher LOS for patients with resistant infections [22, 24]; however, results were not statistically significant. One study inferred the opposite relationship but also with non-statistically significant results [26]. Two studies did not compare results between patients with resistant infections and comparator groups [23, 28], two studies mentioned a prolonged hospitalization without showing the duration [22, 25], and one study did not address the outcome [27]. These studies have reported and compared mean LOS between exposure and comparator groups without estimating the attributable cost due to excess LOS, which contribute to direct and indirect costs.
Antibacterial medication use
Three studies investigated the suitability of the treatments provided to infected patients. The studies assessed the effectiveness of the medications prescribed to treat Acinetobacter baumanii [27], Methicillin-Resistant Staphylococcus Aureus (MRSA) [23], and Extended-Spectrum Beta-Lactamases (ESBL) Escherichia coli [28]. Inappropriate treatment was estimated to have occurred in between 53 and 63% of cases. One of the studies found that MDR infections and inadequate antibiotic treatment were associated with increased mortality [27].
Number of procedures
One study evaluated the average number of procedures per patient [26], which contributes to direct costs. The authors reported that patients with MDR infections needed, on average, slightly more surgeries than patients with non-MDR infections; however, results were not statistically significant.
Health outcomes
Amputation
One study assessed the risk of amputation in the different groups [26], which entails direct, indirect and intangible costs through increased hospitalization, decreased mobility and quality of life, loss of employment, and prosthesis and rehabilitation services usage. Patients with MDR infections had an 83% higher risk of amputation than patients with non-MDR infections; however, results were statistically non-significant (95% CI 0.34–9.89).
Mortality
Three studies reported higher mortality in patients with resistant infections [22, 26, 27]. Two studies did not compare results to comparator groups [22, 25], and three studies did not address the outcome [23, 24, 28]. Studies that assessed the association did not find statistical significance.