Introduction
Campylobacter spp. are one of the leading causes of foodborne illness with an estimated 447 cases per 100 000 persons annually in Canada, and 16% of foodborne illnesses globally [Reference Thomas1, 2]. Campylobacter spp. infections are often self-limiting and are characterised by acute gastroenteritis, commonly diarrhoea, abdominal pain and fever, with severe illness leading to hospitalisation and septicaemia [3]. Common sources of Campylobacter spp. include contaminated food or water, primarily undercooked chicken [3].
Due to the self-limiting nature of campylobacteriosis, antimicrobial treatment is mainly recommended for cases with severe or prolonged infections and immunocompromised patients [Reference Riley4]. Fluoroquinolones are often used for treatment of complicated human Campylobacter spp. infections [Reference Blaser, Engberg, Nachamkin, Szymanski and Blaser5]. The term (fluoro)quinolone-resistant (FQR) used in this review refers to both the quinolones, a group of broad-spectrum bactericidal agents and fluoroquinolones, a class of quinolone-derived antibiotics, which include, but are not limited to ciprofloxacin, levofloxacin and moxifloxacin [6]. Fluoroquinolones are classified as ‘Category I: Very High Importance’ in human medicine by Health Canada's Veterinary Drugs Directorate (VDD) [7]. This classification indicates that fluoroquinolones are one of the limited available therapies for treating serious Campylobacter spp. infections which means resistance to these drugs may leave patients with few or no options for effective treatment [7]. Quinolones (except fluoroquinolones) are classified as ‘Category II: High Importance’ in human medicine by Health Canada's VDD [7]. Quinolones (including fluoroquinlones) are classified as ‘Critically Important Antimicrobials’ used in human medicine by the World Health Organization (WHO) [8]. The WHO listed ‘Campylobacter, fluoroquinolone-resistant’ as ‘High’ priority in the ‘WHO Priority Pathogens List for Research and Development of New Antibiotics’ [9].
Antimicrobial resistance (AMR) in bacteria, including Campylobacter spp., can develop from a genetic mutation within the organism or acquiring existing resistance genes from other organisms due to selection pressures from antimicrobial use [Reference Luangtongkum10]. Studies in Europe and the United States indicate that fluoroquinolone use in animals can select for resistance in human pathogens, particularly Campylobacter jejuni, and the incidence of human FQR Campylobacter spp. infections increased after these drugs were licensed for food animals use [11]. In 2005, the FDA removed approval for fluoroquinolone use in water for poultry based on a risk assessment that found the development of FQR Campylobacter spp. in poultry and its ability to be transferred to people [12]. A study found similar AMR profiles of Campylobacter spp. isolates from broiler chickens in Canadian farms with chicken meat retailed in different regions of Canada [Reference Dramé13]. In Canada, fluoroquinolones are approved for treating respiratory disease in cattle and pigs, but are not approved for use in poultry [14]. With people, animals and the environment serving as reservoirs for the transmission of such resistance genes, infection with FQR Campylobacter spp. is known to spread through the food chain ultimately posing a threat to food safety and human health [11, 14, Reference Larsson15].
Campylobacter jejuni is the most common type of campylobacter bacteria associated with human campylobacteriosis, and other species reported in Canada include C. coli, C. fetus, C. lari and C. upsaliensis [16]. In a region of Western Canada, a study found that most infections were linked to C. jejuni subtypes associated with cattle, and many occurrences of pathogenic campylobacter species were not detected by the conventional laboratory methodology [Reference Inglis and Boras17]. An additional study looking at campylobacteriosis in this region found that most FQR C. jejuni subtypes were endemic in Canada and mostly linked to cattle and chicken reservoirs [Reference Inglis, Taboada and Boras18]. These studies highlight the need to examine campylobacter at the species level to gain further understanding of reservoirs and transmission mechanisms. The utility of techniques in molecular epidemiology, including whole genome sequencing, will evolve research and enhance our ability to discriminate between strains and advance our assessment of impacts to health [Reference Inglis and Boras17, Reference Inglis, Taboada and Boras18].
Increases in FQR Campylobacter spp. have been observed in Canada and worldwide [Reference Luangtongkum10, Reference Inglis, Taboada and Boras18, 19]. An increase in human FQR Campylobacter spp. isolates was observed in Canada between 2011 and 2018 [19]. This is concerning as the Council of Canadian Academies Expert Panel on the Potential Socio-Economic Impacts of AMR in Canada found that in 2018 AMR cost the Canadian healthcare system an estimated $1.4 billion [20]. Moreover, based on estimates, the Panel reported that resistant infections have a higher mortality rate compared to susceptible infections, likely due to complications from delays in obtaining effective treatment [20]. Burden of illness (BOI) is defined broadly as the direct and indirect impacts or costs of illness which can have clinical outcome and resource implications from the patient, healthcare system and economic perspective [Reference Naylor21, 22]. To estimate the overall human health impact of FQR Campylobacter spp. infections, studies utilise a variety of BOI measures including duration of illness, treatment failure, absenteeism and direct/indirect healthcare costs [11, Reference McDonald23, Reference Mutter, Rosko and Wong24].
A scoping review was undertaken to synthesise available literature. Scoping reviews are a key tool to summarise research findings and knowledge gaps to inform knowledge users, including policy makers and practitioners [Reference Arksey and O'Malley25]. Currently, there is no publication that has synthesised BOI measures related to human FQR Campylobacter spp. infections that the authors are aware of. This synthesis of information will add to existing knowledge and subsequently aid in the development of risk assessments and risk reduction measures related to human FQR Campylobacter spp. infections. The purpose of this review is to summarise how FQR Campylobacter spp. infections are characterised in the literature by (i) describing how BOI associated with FQR is measured and reported; (ii) describing the variability in reporting of study characteristics; and (iii) providing a narrative synthesis of literature that compared BOI measures between FQR Campylobacter spp. infections and those with susceptible infections.
Methods
Literature search
This scoping review was guided by methodology from ‘Preferred Reporting Items for Systematic review and Meta-Analyses extension for Scoping Reviews’ (PRISMA-ScR) [Reference Tricco26]. A protocol was established prior to starting the scoping review and updated prior to the second search (available upon request). A literature search was conducted to capture published BOI data on human infection with FQR Campylobacter spp. The terms included in the search string related to the bacteria of interest (campylobacter), and the exposure terms ((fluoro)quinolone resistance), which were compiled by the research team using the WHO ATC/DDD Index (Table 1) [27]. Terms related to specific campylobacter species were not included to capture all possible camplylobacter infections regardless if they were speciated or if the species was not mentioned. Terms related to the outcome (BOI) were not included to capture all possible BOI measures. Seven databases were searched on 6 February 2020: MEDLINE® via Ovid, EMBASE® via Ovid, Core Collection in Web of Science, Scopus, CAB Abstracts via Ovid, AGRICOLA™ via Ovid and Global Health via Ovid. Grey literature sources included the WHO's Global Index Medicus and Google Scholar (only first 250 results based on relevance). The primary and grey literature searches were repeated on 7 June 2021 and 10 June 2021, respectively (Table S1), both with a publication date limit of 2020 to present. In the updated search, MEDLINE® via Ovid and EMBASE® via Ovid were run without the human filter and with a publication date filter from 2018-present. This was done to ensure that all articles were captured even if there was a delay in indexing (time interval between entry into database and indexing) [Reference Irwin and Rackham28].
Search string of for primary search of Scopus (February 2020) used to identify literature associated with (fluoro)quinolone-resistant Campylobacter spp. infections
Inclusion criteria
Articles in this review were included based on the following criteria: population, exposure/comparator, outcome, study design and language. To be included, studies needed to be about people of any age with a laboratory confirmed infection with any FQR-Campylobacter spp. The comparator group, when appropriate to the study design, included cases with Campylobacter spp. infections that were either susceptible to (fluoro)quinolones or pan-susceptible. The studies must have also evaluated one or more BOI measures that included, but were not limited to, treatment failure, mortality and length of hospital stay (LOS). Lastly, the included studies had to be primary observational studies [Reference Dohoo, Martin and Stryhn29], including theses and dissertation, and published in English or French.
Relevance screening
All citations were imported and de-duplicated in RefWorks© (ProQuest, Ann Arbor, United States) and DistillerSR® (Evidence Partners, Ottawa, Canada). After removal of duplicates, primary screening of titles and abstracts was performed using the inclusion criteria (Table S2). An accelerated screening approach was utilised where articles included by a single reviewer proceeded to secondary screening and agreement by two reviewers was required for excluded references. Secondary screening of full-text was completed by two reviewers independently with inclusion criteria reapplied (Table S3). Conflicts were discussed and resolved between reviewers and if necessary, arbitrated by a third researcher with inclusion criteria applied.
Data extraction
Data extraction was performed by two reviewers individually in DistillerSR®, and conflicts were managed the same as during screening. Descriptive data that were extracted included study characteristics (year of publication, year(s) of data collection, study design, country, study objective, study site(s), campylobacter species), study participant characteristics (underlying common disease process and/or common characteristics, immune status, definition of resistant infections, numbers and selection of cases and comparators, age and sex, type of infections, testing methods and criteria) and BOI measures reported (definitions and results) (Table S4). BOI measures were mapped and qualitatively assessed according to three perspectives outlined in a systematic review by Naylor et al. [Reference Naylor21]. These perspectives included patient perspective (e.g., mortality and morbidity measures), healthcare system perspective, which looks at the burden to healthcare providers (e.g. LOS and hospital costs), and economic perspective, which looks at the impact on the labour force and society (e.g., absenteeism) [Reference Naylor21]. All of these data were synthesised to provide a qualitative summary of BOI measures in the literature that related to human FQR Campylobacter spp. infections.
Results
General
The original and updated literature searches retrieved 3325 (de-duplicated) potentially relevant articles, out of these 26 articles met inclusion criteria and were identified for data extraction (Fig. 1). The data from included articles were collected from 1985 to 2018 and published from 1994 to 2021 (Fig. 2). The most common author reported study designs were case reports (n = 11) and case-comparisons (n = 4). The studies were conducted in 15 different countries with the United States (n = 5) and Japan (n = 3) as the most common (Tables 2 and 3). In terms of Campylobacter spp, the most commonly reported were C. jejuni (n = 16) and Campylobacter coli (n = 7), and six studies did not report the specific species (Tables 2 and 3). Non-case reports (n = 15) were mostly conducted in multi-sites (n = 9) and all case reports were conducted in a single site (Tables 2 and 3). Most studies reported underlying common disease process and/or common characteristics (n = 19), including various comorbidities such as cancer (n = 5), human immunodefiency virus (HIV) (n = 2) and diabetes (n = 2) (Tables 2 and 3). The most common types of infection reported were gastrointestinal (GI) (n = 22) and blood-stream infection (BSI) (n = 8) (Tables 2 and 3). The number of cases with resistant infections ranged from 1 to 1078 and the number of cases with susceptible infections ranged from 3 to 3726 (Table 2). Cases and comparators were identified using various methods, including hospital medical records, patient registries, institutional databases and questionnaire responses.
Literature search and screening process for a scoping review of BOI measures related to human (fluoro)quinolone-resistant Campylobacter spp. infections (adapted from [Reference Tricco26]).
Years of data collection and publication period for data included in the scoping review of BOI measures related to human (fluoro)quinolone-resistant Campylobacter spp. infections. *Frequency of Data Collection: each year of data during a multi-year collection period is recorded as one value.
Characteristics of references (excluding case reports) included in scoping review of BOI measures related to human (fluoro)quinolone-resistant Campylobacter spp. infections
BSI, Bloodstream infection; CIP-R, Ciprofloxacin-resistant; CIP-S, Ciprofloxacin-susceptible; GI, Gastrointestinal; HIV, Human immunodeficiency virus; MSM, Men sexually active with men; NARMS, National Antimicrobial Resistance Monitoring System for Enteric Bacteria; NR, Not reported; N/A. Not available.
Characteristics of case reports included in scoping review of BOI measures related to human (fluoro)quinolone-resistant Campylobacter spp. infections
BSI, Bloodstream infection; GI, Gastrointestinal; UTI, Urinary tract infection.
Case reports detailed the age and sex of participants, where age ranged from 9 months to 78 years and they included six females and eight males (Table 3). Age and sex characteristics were reported by 13 non-case report studies (Table S5). Studies defined cases with FQR infections as being resistant to fluoroquinolones (n = 17), quinolones (n = 4) or both (n = 5) (Table S6). They described the antimicrobial susceptibility testing method, where the most common method was disk diffusion (n = 8), and the most common interpretive criteria used were the Clinical and Laboratory Standards Institute clinical breakpoints (n = 8) (Table S6). Studies reported 26 different BOI measures, which related to the BOI perspective of the patient (n = 23), healthcare system (n = 2) or economy (n = 1). The most commonly identified BOI measures across all studies were hospitalisation, diarrhoea, fever and duration of illness (Table 4).
Identified BOI measures for human (fluoro)quinolone-resistant Campylobacter spp. infections (n = 26 references)
Patient perspective BOI measures
Diarrhoea
Diarrhoea was the most commonly reported patient perspective BOI measure (n = 17). There were several definitions of diarrhoea used including having three or more loose stools in a 24 h period (n = 1), two or more loose stools in a 24 h period with one or more associated GI symptom (n = 1), chronic diarrhoea (three or more bowel movements per day lasting more than two days) (n = 1) and frequency of diarrhoea (n = 1). The remaining 13 studies did not provide a definition. Two studies presented comparisons between FQR and susceptible cases. One study found no difference between the two groups [Reference Evans30], and one study reported diarrhoea in HIV positive and HIV negative patients for both FQR and susceptible cases [Reference Tee and Mijch31] (Table S7). The remaining studies did not provide a comparison between FQR and susceptible infections and only mentioned participants reporting diarrhoea (n = 12), or all participants enroled in the study had diarrhoea (n = 2).
Fever
Fever was the second most commonly reported patient perspective BOI measure (n = 15). A fever was defined as having a temperature of ≥38 °C (n = 4), or no definition was provided (n = 11). Four studies presented comparisons between FQR and susceptible cases. One study found that FQR cases were less likely to report fever than cases with susceptible infections [Reference Evans30] (Table S7). One study reported fever in HIV positive and HIV negative patients for both FQR and susceptible cases [Reference Tee and Mijch31], and two studies found no difference in fever between the FQR and susceptible cases [Reference Unicomb32, Reference Nelson33]. The remaining studies did not provide a comparison between FQR and susceptible infections and only mentioned participants reporting a fever (n = 11).
Duration of illness
Duration of illness was the third most commonly reported patient perspective BOI measure (n = 12). Duration of illness was defined as duration of diarrhoea in days and/or months (n = 7), duration of diarrhoea and fever in days (n = 1), duration of diarrhoea and chest pain in days (n = 1), or no definition was provided (n = 3). Eight studies presented comparisons between FQR and susceptible cases. Three studies found a longer duration of illness in resistant cases compared to susceptible cases, and two of the results were statistically significant [Reference Nelson33–Reference Engberg35] (Table S7). One study reported duration of diarrhoea and fever in HIV positive and HIV negative patients for both FQR and susceptible cases [Reference Tee and Mijch31], and four studies found no difference in duration of illness between the FQR and susceptible cases [Reference Evans30, Reference Unicomb32, Reference Engberg35, 36] (Table S7). The remaining studies did not provide a comparison between FQR and susceptible infections and only mentioned participants reporting duration of illness (n = 5).
Bloody stool
Bloody stool was a reported patient perspective BOI measure by eight studies. Bloody stool was defined as blood found in a stool specimen (n = 1), bloody diarrhoea (n = 1), or no definition was provided (n = 6). Five studies presented comparisons between FQR and susceptible cases. One study reported a higher percentage of FQR cases having bloody diarrhoea compared with susceptible cases [Reference Gupta37] (Table S7). In contrast, one study found that susceptible isolates seemed to cause bloody stools more than FQR isolates [Reference Feodoroff38] (Table S7). Three studies found no difference between FQR and susceptible cases [Reference Evans30, Reference Unicomb32, Reference Nelson33] (Table S7). The remaining studies only mentioned participants reporting bloody stool (n = 3).
Bacteremia
Bacteremia or BSI was a reported patient perspective BOI measure by eight studies. Bacteremia was defined as a positive blood culture for Campylobacter spp. (n = 7), or no definition was provided (n = 1). One study presented bacteremia in both FQR and susceptible cases, however due to the study design all participants had bacteremia [Reference Tee and Mijch31]. The remaining studies only mentioned participants reporting bacteremia irrespective of susceptibility (n = 7).
Abdominal pain
Abdominal pain was a reported patient perspective BOI measure by six studies. There were no definitions provided. Two studies presented comparisons between FQR and susceptible cases and both studies found no difference between the two groups [Reference Evans30, Reference Nelson33] (Table S7). The remaining studies only mentioned participants who reported abdominal pain irrespective of susceptibility (n = 4).
Treatment failure
Treatment failure was a reported patient perspective BOI measure by five studies (Table S7). Treatment failure was defined as lack of resolution of diarrhoea with standard antibiotic treatment (n = 1), persistence of diarrhoea, bacteremia and femoral pain (n = 1), or no definition was provided (n = 3). One study compared treatment failure between FQR and susceptible infections and found higher treatment failure among the FQR cases [Reference Sanders39] (Table S7). The remaining studies only described treatment failure irrespective of susceptibility (n = 4).
Vomiting
Vomiting was a reported patient perspective BOI measure by four studies. There were no definitions provided. Three studies presented comparisons between FQR and susceptible cases and all found no difference between the two groups [Reference Evans30, Reference Unicomb32, Reference Nelson33] (Table S7). The remaining study only mentioned participants who reported vomiting.
Complications
Complications were a reported patient perspective BOI measure by four studies (Table S7). There were no definitions provided and no comparisons between FQR and susceptible infections were made.
Mortality
Mortality was a reported patient perspective BOI measure by three studies. Mortality was defined as death (n = 2), or death during follow-up (n = 1). Two studies compared mortality between FQR and susceptible cases, where one study reported cases of death in HIV positive and HIV negative patients for both FQR and susceptible cases [Reference Tee and Mijch31], and the other study reported the number of deaths in both FQR-resistant and susceptible groups [Reference Molina40] (Table S7). The remaining study did not provide a comparison between FQR and susceptible infections.
Nausea
Nausea was a reported patient perspective BOI measure by three studies. There were no definitions provided. One study presented comparisons between FQR and susceptible cases and found no difference between the two groups [Reference Evans30] (Table S7). The remaining studies only mentioned participants who reported nausea irrespective of susceptibility (n = 2).
Other
There were additional BOI measures under the patient perspective category documented by one included study (Table 4). Two studies presented comparisons between FQR and susceptible cases. One study found that patients with FQR infection had an increased risk of adverse events (a diagnosis of either invasive illness or death) compared with patients with susceptible strains [Reference Helms41] (Table S7). One study reported comparisons between FQR and susceptible cases with regard to muscle aches, painful joints, weakness and headache, but found no significant difference between these groups for any of these outcomes [Reference Evans30] (Table S7). The remaining BOI measures, reduced appetite, night sweats, severe polar, progressive pain in upper extremities due to osteomyelitis, ileocolitis, pharyngotonsillitis and urinary tract infection, were observed in five case reports [Reference Martora42–Reference Pascual46] (Table S7).
Healthcare system perspective BOI measures
Hospitalisation
Hospitalisation was the most commonly reported healthcare system perspective BOI measure (n = 20). Hospitalisation was defined as admission to the hospital (n = 11), requiring hospitalisation (n = 2), presence of a hospital medical record (n = 1) or no definition was provided (n = 6). Seven studies presented comparisons between FQR and susceptible cases. Two studies reported that patients with FQR infections were more likely to be hospitalised compared to patients with susceptible infections, where one study looked at non-travel cases [Reference Gupta37, Reference Ricotta47] (Table S7). Two studies found that hospitalisation was higher with susceptible cases compared to FQR cases, where one study looked at international travel cases [Reference Feodoroff38, Reference Ricotta47] (Table S7). Four studies found no difference between FQR and susceptible infections [Reference Evans30, Reference Unicomb32, Reference Nelson33, 36] (Table S7). The remaining studies only mentioned participants being hospitalised irrespective of susceptibility (n = 13).
Length of hospital stay
LOS was the only other reported healthcare system perspective BOI measure (n = 7). It was defined as LOS in days (n = 2), days from admission to discharge (n = 2), time from admission to death (n = 1) or no definition was provided (n = 2). Two studies presented comparisons between FQR and susceptible cases. One study found that the mean LOS was longer for the susceptible group than for the FQR group [Reference Nelson33] (Table S7). However, one study did not find a difference between median LOS in FQR and susceptible cases [Reference Unicomb32] (Table S7). The remaining studies only mentioned LOS irrespective of susceptibility (n = 4).
Economic perspective BOI measures
Absenteeism
Absenteeism was the only reported economic perspective BOI measure (n = 2). Absenteeism was defined as missed or time off work or school (n = 1) and missed usual activities in days (n = 1). Both of the studies presented comparisons between FQR and susceptible cases and found no difference between both groups [Reference Evans30, Reference Nelson33] (Table S7).
Discussion
This scoping review summarised literature that reported BOI measures associated with human FQR Campylobacter spp. infections three categories (patient, healthcare system and economic perspectives [Reference Naylor21]). The majority of studies reported patient perspective BOI measures, and we found a limited amount of information in the healthcare system and economic BOI perspectives. There were no studies that addressed healthcare costs, which would be useful to gain further insight since there were many hospitalisations associated with FQR Campylobacter spp. infections.
While our scoping review was not specifically designed to evaluate comorbidities, several studies provided insight into BOI due to FQR Campylobacter spp. infection in patients with underlying health conditions, and the most commonly reported were cancer, HIV and diabetes. This knowledge can be important in the clinical care of patients, and may also provide an opportunity for future research into identification and exploration of potential high risk populations in relation to patients with FQR Campylobacter spp. infections.
Similar to Campylobacter spp., there has been research for which BOI related to resistance has been explored in other enteric pathogens. Salmonella spp. are another common cause of foodborne illness that result in gastroenteritis in Canada and globally [Reference Martin48]. A Canadian study found an increased risk of hospitalisation for cases with multidrug-resistant Salmonella enterica serotype Typhimurium compared to cases with susceptible infections [Reference Martin48]. In a Danish study, patients infected with quinolone-resistant S. enterica serotype Typhimurium had a higher risk of invasive illness or death within 90 days of infection than patients infected with pansusceptible strains [Reference Helms, Simonsen and Mølbak49]. In the United States, a study found that BSI occurred more frequently among patients infected with drug resistant non-typhoidal Salmonella than patients with pan-susceptible infection [Reference Varma50]. We are not aware of a published manuscript synthesising BOI measures relating to AMR in non-typhoidal Salmonella and this could be an area for future research.
An important question that clinicians, risk assessors and policy makers seek answers to is whether BOI measures increase with FQR Campylobacter infections when compared to those with susceptible infections? Although answering this question is beyond the scope of our review, we can provide insight into the ability to answer this question with a future systematic review and meta-analysis. The most common study design was case reports and due to their design they did not provide data that compared resistant and susceptible cases. Even among the studies that were not case reports (n = 16), only some studies provided data that compared both groups (n = 11) and they all yielded mixed results. Duration of illness (n = 8) and hospitalisation (n = 7) had the most studies that provided data comparing resistant and susceptible cases. It is possible that a systematic review and meta-analysis could be performed on these two BOI outcomes, however, the heterogeneity of the studies would need to be assessed to evaluate whether it is appropriate to combine the results from the studies. As well, risk of bias would need to be assessed to provide the context necessary for interpreting the meta-analysis results. Our scoping review highlights the need for more well designed studies, which could inform future meta-analyses.
Another difficulty is the lack of standard reporting of BOI measures, as all studies employ different methodologies and reporting metrics. Study designs varied from case reports to large cohort studies where definitions of BOI measures varied and many studies did not provide definitions. These same obstacles have been discussed in recent reviews estimating the global health and economic burden of disease due to AMR in many pathogens [Reference Naylor21, Reference Woolhouse51]. These studies state that in order to improve the precision of such estimates and in turn allocate resources and inform policy, broad scale changes in the collection and reporting of BOI measures may be necessary [Reference Naylor21, Reference Woolhouse51].
In addition to the limitations addressed in previous paragraphs, other limitations of this review include the small number of studies included for synthesis, the lack of Canadian literature, and the lack of recent literature comparing resistant and susceptible groups. Only 26 publications met the inclusion criteria, identifying current knowledge gaps regarding human FQR Campylobacter spp. infections. Only one paper included was published in Canada and it reported BOI outcomes including abdominal cramps, bloody stool, diarrhoea, fever and hospitalisation [Reference Gaudreau52]. However, this study did not compare between resistant and susceptible groups, thus more research is needed to understand BOI associated with FQR Campylobacter spp. in the Canadian context. There is a lack of recent literature comparing resistant and susceptible groups, with the most recent studies included that compared the two groups being published in 2009.
In conclusion, several types of BOI measures were reported in the literature to describe the human health impact of FQR Campylobacter spp. infections, however they were inconsistently defined. Despite the common occurrence of this foodborne pathogen, and particularly because of the limited amount of literature comparing BOI for FQR and susceptible Campylobacter spp. infections, this review illustrates the need for further research in which comparable or standardised measures of BOI are reported with harmonised approaches in study methodology. Key information that clinicians and researchers should collect to assess BOI from FQR Campylobacter spp. infections include details on the infection (e.g. sample source, species, method used for antimicrobial susceptibility testing (AST), AST interpretive criteria), patient characteristics including confounders (e.g. age, sex, co-morbidities/underlying disease), and clearly defined BOI measures (e.g. mortality, morbidity, healthcare system and economic impact) stratified by resistant and susceptible cases (Table S8). This information could lead to a better understanding of BOI due to human FQR Campylobacter spp. infections to inform knowledge users including clinicians, policy makers and risk assessors.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S095026882200139X.
Acknowledgements
The authors would like to thank Jaqueline Dubon and Sarah Garland for their efforts in the original literature search and the Public Health Agency of Canada library for help with article procurement.
Financial support
This research was in part supported by the Genomics Research and Development Initiative-Antimicrobial Resistance (GRDI-AMR).
Conflict of interest
None.
Data availability statement
The authors confirm that the data needed to replicate the findings of this study are available within its supplementary materials.
Open access
