A scientometric study on quantitative microbial risk assessment in water quality analysis across 6 years (2016–2021)

Water scarcity is a growing global challenge in the 21st century despite the resource being vital for the survival of humanity (Farhadkhani et al. 2020). According to the WHO-UNICEF (2017) report, more than 844 million persons lack access to safe and clean potable water and about 159 million consume unprocessed water sourced from lakes and streams among other untreated surface waters. This phenomenon is a function of temporal and spatial variation of water resources, which hinders distribution and access to users (Oki & Quiocho 2020). The situation has further been exacerbated by climate change that has altered hydrological and precipitation patterns to make dry areas drier and wet ones wetter (Nyika 2018). Farhadkhani et al. (2020) also noted that water scarcity is exacerbated by drought, rise in urbanization and population growth. In developing countries of low and middle income, which have limited preparedness and high vulnerability to climate change, the water scarcity situation is dire. The trend is attributable to high water demand due to population growth, poor infrastructure to harvest water the resource and manage its demand, as well as the resultant wastewater (Fuhrimann et al. 2016; Nyika & Onyari 2019; Oki & Quiocho 2020). The pressure is even more in urban and suburban areas with no wastewater management systems and therefore, the population resort to release raw effluent to freshwater bodies resulting in a bigger pollution problem on such resources (Mbanga et al. 2020).

Pollution of water resources and the direct consumption of such untreated water results in waterborne diseases. The waterborne illnesses are characterized by diarrhea and spread by viral, protozoan and bacterial pathogens (Shamsollahi et al. 2019). The illnesses are historical and still remain a great challenge since some of the pathological pathways are not understood (Zhang et al. 2019). In 2016 for instance, more than 500 million deaths in sub-Saharan Africa were associated with waterborne diseases caused by pathogens such as thermotolerant coliforms (Shamsollahi et al. 2019), listeria (Gholipour et al. 2020), Entamoeba, Shigella, Aeromonas spp., Cryptosporidium and Escherichia coli (Mutono et al. 2020). In 2016 alone, global deaths affiliated to waterborne diseases totaled to 829,000, and out of this number, more than 360,000 were children under 5 years (WHO-UNICEF 2017). Inefficient and conventional treatment of contaminated water using techniques such as disinfection, sand filtration, coagulation and flocculation, which is commonly applied, cannot efficiently clear pathogenic microbes from the resource (Shamsollahi et al. 2019). To reduce the likelihood of contracting the diseases, it is imperative to screen water for the microbes including viruses, protozoa and bacteria that cause waterborne diseases. However, the pathogenic microbes in water are many and of different kinds, which makes it cost- and time-inefficient to screen all of them (Zhang et al. 2019; Farhadkhani et al. 2020). To alleviate such challenges, the use of microbial source-tracking techniques and fecal indicators has been used to target such disease-causing pathogens in water resources.

Quantitative microbial risk assessment (QMRA) is one of the microbial source-tracking tools that apply dose–response (DR) models for specific disease-causing microbes to assay a variety of exposure scenarios and approximate the affiliated risks to human health (Haas et al. 2014). The tool entails the following four interrelated processes (Haas et al. 2014): (1) identification of hazards, (2) assessment of exposure, (3) assessment of DR and (4) risk characterization. QMRA quantifies the levels of reference pathogens at the exposure points by approximating the sources, fate, transport systems to human and the ingestible volumes to determine the dose. The resultant dose alongside the DR model for each pathogen is used to generate the associated health risks (Federigi et al. 2019). The QMRA approach was first used for potable water to define health targets of the resource's consumption by WHO (2011). It has been used to approximate risks of wastewater and greywater use in irrigation (USEPA 2017; Farhadkhani et al. 2020), sewage sludge ingestion by sewerage plant workers (Gholipour et al. 2020) and recreational waters (Gitter et al. 2020).

Scientometric analysis or bibliometric mapping that intersects in the information science, philology, mathematics and statistics disciplines is helpful in such undertakings. Bibliometric analysis assesses the scientific evolution of research outputs among scientists, journals, regions and institutions on a given topic using performance indicators (Cobo et al. 2015). The approach has been used to study various research fields including vague decision-making (Liu & Liao 2017), economics and business (Merigo et al. 2016), sustainable energy (Hache & Palle 2019) and tropical medicine (Flaherty & Browne 2016).

The potential to use the QMRA tool in water quality assessment is high since its insights can be translated to better actions toward water resources management (Hamilton & Haas 2016). However, the process is yet to be understood exhaustively due to the following reasons. First, there are differences in microbial pathways and modeling uncertainties during pathogen approximation and risk characterization (Petterson & Ashbolt 2016). Second, the interpretation of QMRA results is influenced by the health risks of the pathogen of focus and the modeling assumptions made, which differ significantly (Petterson et al. 2015). Third, the use of QMRA for water analysis in areas with high vulnerability to water pollution such as developing countries of Africa and Asia is not exhaustive and results of findings from different studies are not congruent.

There is a need to understand the trends in research involving the application of QMRA in water quality assessment to identify the institutions and countries that have greatly contributed to the topic as well as research gaps. There is a need to understand the research areas in the publications on QMRA in water quality analysis as they will be useful in setting future research agendas. The evaluation of research trends on QMRA in water quality assessment research and publications to the best of our knowledge has not been conducted in previous studies. This study, therefore, applied scientometric analysis to evaluate the scientific progress on QMRA in water quality analysis in the years 2016–2021, which could be essential to water resources managers and researchers on the directions of the field now and in the future. The study also identified key actors and compared their research activities in reference to the topic (QMRA in water quality assessment) using evidence-based publications. Additionally, the study aimed to perform a detailed scientometric analysis of scientific production in the field of QMRA in water analysis during the 2016–2021 period using various performance indicators such as most active journals, authors, institutions and countries, as well as bibliometric mapping.

The data used in this analysis were collected from the Web of Science (WoS) database and specifically from the arts and humanities citation, conference proceedings citation, emerging sources citation, social sciences citation and science citation index-expanded indices. The search criterion was on topic basis using the search query, ‘quantitative microbial risk analysis and water’ or ‘QMRA and water’ or ‘microbial risk analysis and water’. After the search, the full records of the article journals including their authors, title, abstract, publication date, affiliation and name of the publishing article were downloaded for use as primary data sources for this study. A journal citation report was also downloaded from the database to relate the articles searched to the journals they were published and their affiliated citation references, as well as scientific disciplines. The most cited journals were also related to their associated impact factor.

The most active countries and institutions and their contribution to QMRA In water quality assessment research were assayed based on the VOSviewer program mapping to relate the documents and authors affiliated to them. The VOSviewer software interrelates various aspects of articles using circles and curved lines (Van Eck & Waltman 2010). The latter defines the strength of interrelations of research aspects, while circles represent specific items whose relationship is being determined. Highly ranked items are represented by larger circles and thicker curved lines and vice versa. The circles and lines are of different colors representing clusters based on common features such as similar research areas, co-authorship and similar institutional collaborations. Thicker curved lines are suggestive of strong inter-relationships of the aspects and items in question. Collaborations between and among authors as well as institutions were defined by the author affiliation addresses to establish their co-occurrence (van Haselen 2007).

Document types were limited to only research articles since they provided empirical findings. Review articles were also excluded from this analysis. Furthermore, the search based on years was narrowed between 2016 and 2021 for the following aspects keyword analysis, author analysis based on their institutional or country of origin, article analysis based on co-occurrence and bibliographic coupling, distribution of journals, language used and the disciplines of the documents searched. The yearly limitation was done to get the most relevant and up-to-date information of QMRA in water quality assessment. Keywords by both authors and the database were analyzed based on their co-occurrence and rank, while the Hirsch (h) index was used as a representative indicator of the scientific impact resulting from the searched articles (Hirsch 2007; Bornmann & Daniel 2009). Other qualitative and quantitative indicators of these analyses included the total number of citation counts and the growth trend of publications. They were quantified using the Microsoft Excel tool to rank them based on percentage output.

Using the aforementioned research query, a total of 1,873 documents on QMRA and water quality assessment were found, most of which included research papers. After filtering the year of publication between 2016 and September 2021, the initial search was reduced to 1,115 documents. A further filtering of the results to include only research articles resulted in 1,013 documents. The articles found were not mutually exclusive in the database and they contributed to a majority (90.6%) of the document types. Others included (number, percentage): review articles (99, 8%), early access (24, 2.1%), proceedings papers (16, 1.4%), editorial material (3, 0.3%) and data papers (1, 0.09%).

The selection of journal articles exclusively for this analysis was due to their dominance and by virtue of them undergoing peer-review hence having credible information. Existent bibliometric studies have applied this logic citing that journal articles are more reliable to use owing to thorough scrutiny of their data and information during peer-review processes (Cordoso et al. 2020; Sweileh 2020; Igwaran & Edoamodu 2021; Mbogning et al. 2021). On the basis of language, a majority of the searched articles were published in English (1,007; 99.4%), while other minority languages included Spanish at 0.3% (three articles), German, Polish and Portuguese at 0.1% each (one article). The trend was expected since the English language has an unquestionable role as the lingua franca of research and academic communication globally (Karoly et al. 2020). In the evaluated years, a significant growth in publications on the search query was established, as shown in Figure 1.

The total searched articles on QMRA in water quality assessment in this analysis were published in 320 journals. The record count of articles, impact factor, impact factor rank and h-index of the top 20 most active journals are shown in Table 1. Twenty of the journals had published ≥10 articles on the searched topic. Science of the Total Environment journal published the highest number of articles (112, 11.056%) of the total count and was ranked seventh based on its impact factor, while Water Research was the second most active with 7.5% of all searched documents and was ranked first based on its impact factor. The relationship between the impact factor of the top most active journals and the h-index was also assayed, as shown in Figure 2. Establishing the relationship between the h-index and the impact factor was important because both metrics are standard and common measures of academic performance and quality for authors and publishers (Grech & Rizk 2018). There was no clear trend in the two metrics of journal output as in some cases there was a directly proportional relationship, while in others it was indirect. The observation could be because the impact factor is ranked after every 2 years based on fluctuations in citations of a particular journal, while the h-index indicates the average productivity and citations of a journal and authors over the years (Grech & Rizk 2018).

A bibliographic coupling of the identified journals was done using the VOSviewer program, as shown in Figure 3. The threshold of this analysis was set at five citations for each article and only 37 of the total 320 meet these conditions. Science of the Total Environment, which is a journal that publishes research on all environments including the hydrosphere, and Water Research journal that publishes research on water cycle, quality and management aspects had the largest circles and the thickest connecting lines compared to other journals. This observation inferred to the high citations and article output of the two journals contributing to a high total link strength compared to the others. Other journals that were the top most active publishers, as shown in Table 1, also had considerably stronger link strength due to their high citing publications. Journals that were clustered together with similar colors in the bibliographic network represented similarities in their publishing scope as Gautam (2019) reported in a study applying scientometrics to analyze research publications using classification schemes.

The contributions of different regions and countries to the search query were analyzed based on the research output. Two documents were, however, excluded due to lack of information on the author addresses. A total of 97 countries were involved in research on QMRA in water quality assessment and 7 and 34 out of this total had more than 50 and 10 publication outputs, respectively. The USA and China were the first two most active countries with a contribution of 51.7% of all searched documents. This trend is attributable to the global role of the two regions in promoting network meta-analysis research (Shi et al. 2021). Unlike poor developing countries, the domination by mainstream developed nations of Europe and the USA and fast transitioning developing countries such as China was not a surprising trend as it has been observed in other scientometric studies (Nyika et al. 2021a, 2021b). The pattern was affiliated with greater financial, human and infrastructural capacity to conduct research leading to higher research collaborations locally and internationally (Pesta et al. 2018; Wang et al. 2018; Nyika et al. 2021b). A similar trend was evident from the bibliographic coupling of countries where most developed countries led by the USA and China dominated in research on QMRA in water quality evaluation, as shown in Figure 4. Characteristics of the first 20 most active countries are presented in Table 2. A higher h-index in this case was correlated to higher research output and citation in a particular region. The USA had the highest research output and h-index corresponding to its high potential to conduct research. China, on the other hand, recorded lower h-index than other European countries such as England and Germany as well as Canada and Australia. The trend could be attributable to the inter- and intra-cooperation of European and American countries in research, which is close unlike China whose cooperation tendencies are reserved (Shi et al. 2021).

Using the affiliation of authors, the WoS outlined the institutions involved in QMRA in water quality assessment publication. Institutional analysis in this case supplemented country analysis due to complexities in addresses used by some authors and to avoid errors during article aggregation. According to Noyons et al. (2003), using address data in bibliometric analysis is complex and instead, main organizations such as companies, research institutes and universities are used. Of all the 1,011 articles that had addresses, 1,609 institutional entries were made and 40 of them had at least 10 published articles. The top 20 most active institutes are shown in Table 3. The Chinese Academy of Sciences and the University of California System topped the list of the most active institutions in publication on the search query with records totaling to about 8%. Most of the other top publishing institutes were from institutions of developed countries of Europe and America that have higher capacity to conduct research compared to Asian and African countries as noted in other scientometric studies (Wang et al. 2018; Izuchukwu et al. 2020).

Collaboration among institutions, which is an approach to disseminating and creating knowledge according to Cordoso et al. (2020), was assayed in this research. Ding et al. (2000) supported the suggestion stating that collaborations are concerted scientific efforts to solve impossible global issues. Using the VOSviewer program that mapped the co-authorship of institutions, as shown in Figure 5, collaborative efforts were established. The conditions for the software were a minimum of 10 documents and 50 citations per organization where 23 of all organizations met this threshold. The Chinese Academy of Sciences, US-EPA and Drexel University were some of the best collaborators. Asian countries with exception of China and African institutions did not appear in the top collaborators in the search query. This trend is similar in all scientific fields as observed by Mbogning Fonkou et al. (2021) who suggested that the research contribution of Asian and African continents is significantly less compared to other regions due to inadequate research networks and infrastructure that result in low productivity. Another author attributes collaborative efforts in the developed world to be higher than developing nations owing to the research dependency of the latter (Pouris 2017). The trend also suggests the need for developing countries to mobilize their own limited resources to conduct research through owned initiatives.

Keywords, which are representative of an author's most significant opinions of an article and inform of the trends of a given research topic according to Pesta et al. (2018), were analyzed in this study. The keywords were also used to categorize the research themes of the search query based on their similarity and relationship (Chen et al. 2019; Sweileh 2020). In addition, keywords show knowledge concepts of a search and define their research domain structure according to Chen & Xiao (2016). In this scientometric analysis, keyword analysis showed the most popular subjects covered by the search query. The co-occurrence of the words was assayed using the VOSviewer program where a total of 5,446 keywords were identified. A further thresholding of the words based on 30 times occurrence resulted in 50 words whose mapping is shown in Figure 6. The keywords were clustered into the following four research themes: red, green, yellow and blue groups based on common aspects in the words.

The red cluster had research themes emphasizing the need for microbial risk characterization to identify microbial communities such as bacteria found in environmental samples such as soil and water that can lead to illnesses once ingested or inhaled. The emphasis was limited to risk characterization of water since the VOSviewer analysis was guided by the searched keywords on the WoS database. In addition, most QMRA studies are mainly done on water systems to investigate their risk and plan on mitigation measures Mohamed et al. (2018) highlighted. The blue cluster had research themes defining the QMRA process. Specific words pointed to processes such as fecal indicator assessment of E. coli, microbial source-tracking and risk characterization of contaminated water including ground-, surface- and wastewater. The green cluster had research themes pointing to the microbial communities and pathogens whose infectious nature, associated disease prevalence and exposure levels have been established using the QMRA approach. These included Salmonella, Norovirus and Cryptosporidium species. This observation concurs with Alireza et al. (2015) who supposed that Salmonella and Cryptosporidium species are the most researched microbes using QMRA but adds Norovirus to the list. The yellow cluster had research themes on the factors to consider during the risk characterization. These include the temperature of the water that determines the growth and reproduction of microbial communities, the water quality that influences the pathogenic microbes, which thrive to infectious levels and the ultimate uses of the water evaluated using QMRA. Evidently, QMRA has been widely researched to evaluate contamination in wastewater and drinking water and in particular, to establish the effect of E. coli among other microbial communities in searched articles based on the large circles and thick curved lines of these keywords that were indicative of a high link strength (Figure 6). The research themes pointed to groundwater, drinking water, surface water and wastewater as the significant environmental media associated with microbial pollution. In a related bibliometric analysis, Alireza et al. (2015) established that QMRA was mainly used to evaluate microbial contamination and transmission in drinking water and food. In addition, all themes showed that wastewater is a risk source of microbial pollution since it is a major threat to water quality.

Using the research areas from the WoS database, the most trending subjects related to QMRA in water quality assessment were established. These research areas were related to the most current research themes associated with the search query. From the total searches, 69 research areas were populated and 19 had more than 10 published articles. The 10 top most published research areas are shown in Table 4. Environmental sciences ecology that deals with the dynamics in the interaction of organisms (in this case microbes) with the environment (particularly water resources) was the most published area accounting for more than 60% of total publications. This observation could be because QMRA's role is to link disease-causing microbial agents found in the environment to water safety and water-based illness dynamics for improved pathogen management (Petterson & Ashbolt 2016). The interactions of complex materials such as contaminants and ecology including microbes in various environments such as water resources can be best characterized and modeled using QMRA (Duarte et al. 2019). Other top publishing research areas were engineering, water resources and microbiology owing to the various processes of QMRA, which require the basics of the research areas. Such principles include water treatment recommendations to cleanse its microbial content, the source of polluted water as either surface-, ground- or wastewater and the specific microbe propagating water contamination.

Of the 1,013 articles used for this study, a citation analysis showed that 31,194 authors were involved. All the documents had 8,173 citing articles including self-citations and were cited 9,950 times with a total h-index of 39 and a mean of 9.77 citation times per article. These statistics were, however, subject to change, used only 6 years based on the study yearly limitation and hence not used to indicate the growth in output on the research query. Aksnes et al. (2019) suggested that citations are not accurate indicators in bibliometric analysis as they are controlled by multiple technical factors. Additionally, Shao & Zheng (2016) observed that citations differ based on the articles or documents accumulated in the search engine or database/s used for scientometric analysis. In this study, co-citations were focused on to establish the ability to collaborate among authors. This was done with the logic that co-citations enable the identification of clusters in co-cited pairs, which equips scholars with insights into scientific research structure and knowledge base (Leung et al. 2017). Of the 31,194 total authors populated from the search query, the limit for co-citation analysis in the VOSviewer was set at a minimum of 40 citations per pair of authors and only 38 met this threshold. Of the results, searches without specific authors were filtered out, which changed the research output to 33. The mapping of their interrelationship is shown in Figure 7. Similar clustering based on coloring indicated co-citations from similar institutions or regions.

Charles Haas, a professor of Environmental Engineering at Drexel University, topped in the collaborations with 272 co-citations. He has published extensively on QMRA in water quality assessment (Haas et al. 2014; Hamilton & Haas 2016; Hamilton et al. 2017). Jeffrey Soller, who is a lead investigator in the Soller Environmental, LLC, was the second-best collaborator with 164 citations. He has several publications on the research query (Soller et al. 2014; Ichida et al. 2016; Boehm & Soller 2020). Other prominent authors in the searched topic included Peter Teunis, Nicholas Ashbolt and Mary Schoen. The authors have several publications on microbial risk characterization in water systems using the QMRA approach.

The aim of this study was to analyze the publication trends on the QMRA use in water quality assessment between 2016 and 2021. When using scientometric analysis some important factors are considered to alleviate bias of bibliometrics on recently published articles. According to FaisalUddin et al. (2004), such prejudice results in the exclusion of some vital publications in the analysis for being deemed obsolete. This scientometric analysis established trends and gaps on the search query to inform of its future prospects and direction. Evidently, the analysis showed that publications on QMRA in water quality analyses were found in high-impact journals such as the Water Research and Science of the Total Environment. Additionally, the most active countries in publishing on the search query had a high h-index. The total number of publications, impact factor of the journals they were published and their citations in this case were used as focal criteria to define the impact of the resultant articles and measure their academic impact as Sobhy (2016) advised.

The country and institutional analysis of publication activity on the search query showed a high output for developed countries compared to developing ones. European countries, China and USA showed high productivity, which is associated with their high economic potential, research infrastructure and high ability to collaborate with others as Igwaran & Edoamodu (2021) established in a study using bibliometric analyses. High publication output could also be affiliated to great awareness on the need to conserve the quality of water and integrity of water resources considering its finite nature and the deteriorating quality, quantity and safety of such resources in an era of climate change and increased pollution globally (Boehm & Soller 2020).

Collaborations among scientists are foundations of existent research in all scientific disciplines. In collaboration networks, scientists are seeking for solutions and answers on socio-economic issues whose nature is multidisciplinary as Sonnenwald (2007) noted. In public health issues such as the QMRA use in water quality assessment, the analysis of collaborative networks is imperative owing to the multi-stakeholder involvement and its inter-disciplinary nature. In this study, collaborations on the search query were mainly from authors drawn from institutions in developed countries. While China had high quantity of research output, its collaboration with institutions outside China was limited. Overall, there was an exponential increase in publications on QMRA in water quality assessment during the evaluated period. The trend is correlated to the productivity of involved authors and the entire scientific community as Merigo & Yang (2017) deduced. Countries such as the USA and China had high research activity regarding the search query based on their research output. A number of institutions drawn from developed countries were also actively involved in research on the search query based on the VOSviewer mapping analysis. However, there is a need to extend research inputs and collaborations to poor developing countries of Africa and Asia to enhance their research activity and output for a holistic contribution to knowledge. This suggestion has been made in other bibliometric studies (Cordoso et al. 2020; Izuchukwu et al. 2020; Nyika et al. 2021a).

The study being the only scientometric analysis on QMRA in water quality assessment to the best of our knowledge, generated comprehensive and broad list of journals, countries, institutions and authors involved in related research alongside the trend in publication growth. In addition to this, the study applied the keywords and research areas analyses to establish the most prolific research themes and topics regarding the search query. Despite these strengths, several limitations of the study were pointed out. First, the classification of keywords into themes could be subject to bias and hence the need to advance research for a better classification system that quantifies and qualifies research trends and patterns. The exclusive use of articles sourced from the WoS database though the most recommended for scientometrics does not reflect the accurate state of research on the search query. Future studies, therefore, should focus on expanding the search to other databases such as Scopus to enable a holistic search. Although standardization of keywords, institutions and authors was done, reported results could be erroneous due to keywords that have analogous meaning but varied expressions, common institutions with different names and an author with more than one affiliation in one publication.

This research provided a qualitative and quantitative analysis of the most active journals, authors, countries and institutions on QMRA in water quality analyses between 2016 and 2021. In this scientometric analysis, the VOSviewer software was used in mapping the various indicators of research output. The analyses provided insights into the research activities of various authors and the collaborations of various institutions and countries. High-impact journals such as Science of the Total Environment and Water Research were actively involved in publishing on the search query with an 8% throughput of total articles. The USA and China were the most active regions in research on the topic with 51.7% of total searched results. Institutions located in the regions equally had high research output. However, it was observed that developing countries were less active in research activities and the trend was attributable to lack of finances and infrastructure to conduct research as well as limited collaborative networks with developed countries. It is, therefore, recommendable to involve such poor countries by offering research funds, incentives and exchange programs among educational institutions to enable knowledge sharing and enhancement globally. Collaborations with countries such as China could also enhance a multidisciplinary approach to solving water quality issues and demystifying the uncertainties in the QMRA process. The methodology used in this research also needs reevaluation and improvement to expand the search to include other databases apart from the WoS, to use other quantitative analysis methods for various research theme identification of the search query and to avoid bias during bibliographic coupling.

All relevant data are included in the paper or its Supplementary Information.