Citation Analysis May Severely Underestimate the Impact of Clinical Research as Compared to Basic Research

Background

Citation analysis has become an important tool for research performance assessment in the medical sciences. However, different areas of medical research may have considerably different citation practices, even within the same medical field. Because of this, it is unclear to what extent citation-based bibliometric indicators allow for valid comparisons between research units active in different areas of medical research.

Methodology

A visualization methodology is introduced that reveals differences in citation practices between medical research areas. The methodology extracts terms from the titles and abstracts of a large collection of publications and uses these terms to visualize the structure of a medical field and to indicate how research areas within this field differ from each other in their average citation impact.

Results

Visualizations are provided for 32 medical fields, defined based on journal subject categories in the Web of Science database. The analysis focuses on three fields: Cardiac & cardiovascular systems, Clinical neurology, and Surgery. In each of these fields, there turn out to be large differences in citation practices between research areas. Low-impact research areas tend to focus on clinical intervention research, while high-impact research areas are often more oriented on basic and diagnostic research.

Conclusions

Popular bibliometric indicators, such as the h-index and the impact factor, do not correct for differences in citation practices between medical fields. These indicators therefore cannot be used to make accurate between-field comparisons. More sophisticated bibliometric indicators do correct for field differences but still fail to take into account within-field heterogeneity in citation practices. As a consequence, the citation impact of clinical intervention research may be substantially underestimated in comparison with basic and diagnostic research.     

Network Effects on Scientific Collaborations

Background

The analysis of co-authorship network aims at exploring the impact of network structure on the outcome of scientific collaborations and research publications. However, little is known about what network properties are associated with authors who have increased number of joint publications and are being cited highly.

Methodology/Principal Findings

Measures of social network analysis, for example network centrality and tie strength, have been utilized extensively in current co-authorship literature to explore different behavioural patterns of co-authorship networks. Using three SNA measures (i.e., degree centrality, closeness centrality and betweenness centrality), we explore scientific collaboration networks to understand factors influencing performance (i.e., citation count) and formation (tie strength between authors) of such networks. A citation count is the number of times an article is cited by other articles. We use co-authorship dataset of the research field of ‘steel structure’ for the year 2005 to 2009. To measure the strength of scientific collaboration between two authors, we consider the number of articles co-authored by them. In this study, we examine how citation count of a scientific publication is influenced by different centrality measures of its co-author(s) in a co-authorship network. We further analyze the impact of the network positions of authors on the strength of their scientific collaborations. We use both correlation and regression methods for data analysis leading to statistical validation. We identify that citation count of a research article is positively correlated with the degree centrality and betweenness centrality values of its co-author(s). Also, we reveal that degree centrality and betweenness centrality values of authors in a co-authorship network are positively correlated with the strength of their scientific collaborations.

Conclusions/Significance

Authors’ network positions in co-authorship networks influence the performance (i.e., citation count) and formation (i.e., tie strength) of scientific collaborations.     

Do Scientific Advancements Lean on the Shoulders of Giants? A Bibliometric Investigation of the Ortega Hypothesis

Background

In contrast to Newton''s well-known aphorism that he had been able “to see further only by standing on the shoulders of giants,” one attributes to the Spanish philosopher Ortega y Gasset the hypothesis saying that top-level research cannot be successful without a mass of medium researchers on which the top rests comparable to an iceberg.

Methodology/Principal Findings

The Ortega hypothesis predicts that highly-cited papers and medium-cited (or lowly-cited) papers would equally refer to papers with a medium impact. The Newton hypothesis would be supported if the top-level research more frequently cites previously highly-cited work than that medium-level research cites highly-cited work. Our analysis is based on (i) all articles and proceedings papers which were published in 2003 in the life sciences, health sciences, physical sciences, and social sciences, and (ii) all articles and proceeding papers which were cited within these publications. The results show that highly-cited work in all scientific fields more frequently cites previously highly-cited papers than that medium-cited work cites highly-cited work.

Conclusions/Significance

We demonstrate that papers contributing to the scientific progress in a field lean to a larger extent on previously important contributions than papers contributing little. These findings support the Newton hypothesis and call into question the Ortega hypothesis (given our usage of citation counts as a proxy for impact).     

Fifty-Year Fate and Impact of General Medical Journals

Background

Influential medical journals shape medical science and practice and their prestige is usually appraised by citation impact metrics, such as the journal impact factor. However, how permanent are medical journals and how stable is their impact over time?

Methods and Results

We evaluated what happened to general medical journals that were publishing papers half a century ago, in 1959. Data were retrieved from ISI Web of Science for citations and PubMed (Journals function) for journal history. Of 27 eligible journals publishing in 1959, 4 have stopped circulation (including two of the most prestigious journals in 1959) and another 7 changed name between 1959 and 2009. Only 6 of these 27 journals have been published continuously with their initial name since they started circulation. The citation impact of papers published in 1959 gives a very different picture from the current journal impact factor; the correlation between the two is non-significant and very close to zero. Only 13 of the 5,223 papers published in 1959 received at least 5 citations in 2009.

Conclusions

Journals are more permanent entities than single papers, but they are also subject to major change and their relative prominence can change markedly over time.     

The Impact of Boundary Spanning Scholarly Publications and Patents

Background

Human knowledge and innovation are recorded in two media: scholarly publication and patents. These records not only document a new scientific insight or new method developed, but they also carefully cite prior work upon which the innovation is built.

Methodology

We quantify the impact of information flow across fields using two large citation dataset: one spanning over a century of scholarly work in the natural sciences, social sciences and humanities, and second spanning a quarter century of United States patents.

Conclusions

We find that a publication''s citing across disciplines is tied to its subsequent impact. In the case of patents and natural science publications, those that are cited at least once are cited slightly more when they draw on research outside of their area. In contrast, in the social sciences, citing within one''s own field tends to be positively correlated with impact.     

Does Collocation Inform the Impact of Collaboration?

Background

It has been shown that large interdisciplinary teams working across geography are more likely to be impactful. We asked whether the physical proximity of collaborators remained a strong predictor of the scientific impact of their research as measured by citations of the resulting publications.

Methodology/Principal Findings

Articles published by Harvard investigators from 1993 to 2003 with at least two authors were identified in the domain of biomedical science. Each collaboration was geocoded to the precise three-dimensional location of its authors. Physical distances between any two coauthors were calculated and associated with corresponding citations. Relationship between distance of coauthors and citations for four author relationships (first-last, first-middle, last-middle, and middle-middle) were investigated at different spatial scales. At all sizes of collaborations (from two authors to dozens of authors), geographical proximity between first and last author is highly informative of impact at the microscale (i.e. within building) and beyond. The mean citation for first-last author relationship decreased as the distance between them increased in less than one km range as well as in the three categorized ranges (in the same building, same city, or different city). Such a trend was not seen in other three author relationships.

Conclusions/Significance

Despite the positive impact of emerging communication technologies on scientific research, our results provide striking evidence for the role of physical proximity as a predictor of the impact of collaborations.     

Looking for Landmarks: The Role of Expert Review and Bibliometric Analysis in Evaluating Scientific Publication Outputs

Objective

To compare expert assessment with bibliometric indicators as tools to assess the quality and importance of scientific research papers.

Methods and Materials

Shortly after their publication in 2005, the quality and importance of a cohort of nearly 700 Wellcome Trust (WT) associated research papers were assessed by expert reviewers; each paper was reviewed by two WT expert reviewers. After 3 years, we compared this initial assessment with other measures of paper impact.

Results

Shortly after publication, 62 (9%) of the 687 research papers were determined to describe at least a ‘major addition to knowledge’ –6 were thought to be ‘landmark’ papers. At an aggregate level, after 3 years, there was a strong positive association between expert assessment and impact as measured by number of citations and F1000 rating. However, there were some important exceptions indicating that bibliometric measures may not be sufficient in isolation as measures of research quality and importance, and especially not for assessing single papers or small groups of research publications.

Conclusion

When attempting to assess the quality and importance of research papers, we found that sole reliance on bibliometric indicators would have led us to miss papers containing important results as judged by expert review. In particular, some papers that were highly rated by experts were not highly cited during the first three years after publication. Tools that link expert peer reviews of research paper quality and importance to more quantitative indicators, such as citation analysis would be valuable additions to the field of research assessment and evaluation.     

US News and World Report Cancer Hospital Rankings: Do They Reflect Measures of Research Productivity?

Context

Prior research has faulted the US News and World Report hospital specialty rankings for excessive reliance on reputation, a subjective measure of a hospital''s performance.

Objective

To determine whether and to what extent reputation correlates with objective measures of research productivity among cancer hospitals.

Design

A retrospective observational study.

Setting

Automated search of NIH Reporter, BioEntrez, BioMedline and Clinicaltrials.gov databases.

Participants

The 50 highest ranked cancer hospitals in 2013''s US News and World Report Rankings.

Exposure

We ascertained the number of NCI funded grants, and the cumulative funds received by each cancer center. Additionally, we identified the number of phase I, phase II, and phase III studies published and indexed in MEDLINE, and registered at clinicaltrials.gov. All counts were over the preceding 5 years. For published articles, we summed the impact factor of the journals in which they appeared. Trials were attributed to centers on the basis of the affiliation of the lead author or study principal investigator.

Main Outcome

Correlation coefficients from simple and multiple linear regressions for measures of research productivity and a center''s reputation.

Results

All measures of research productivity demonstrated robust correlation with reputation (mean r-squared  = 0.65, median r-squared = 0.68, minimum r-squared = .41, maximum r-squared = 0.80). A multivariable model showed that 93% of the variation in reputation is explained by objective measures.

Conclusion

Contrary to prior criticism, the majority of reputation, used in US News and World Rankings, can be explained by objective measures of research productivity among cancer hospitals.     

Patients' and observers' perceptions of involvement differ. Validation study on inter-relating measures for shared decision making

Objective

Patient involvement into medical decisions as conceived in the shared decision making method (SDM) is essential in evidence based medicine. However, it is not conclusively evident how best to define, realize and evaluate involvement to enable patients making informed choices. We aimed at investigating the ability of four measures to indicate patient involvement. While use and reporting of these instruments might imply wide overlap regarding the addressed constructs this assumption seems questionable with respect to the diversity of the perspectives from which the assessments are administered.

Methods

The study investigated a nested cohort (N = 79) of a randomized trial evaluating a patient decision aid on immunotherapy for multiple sclerosis. Convergent validities were calculated between observer ratings of videotaped physician-patient consultations (OPTION) and patients'' perceptions of the communication (Shared Decision Making Questionnaire, Control Preference Scale & Decisional Conflict Scale).

Results

OPTION reliability was high to excellent. Communication performance was low according to OPTION and high according to the three patient administered measures. No correlations were found between observer and patient judges, neither for means nor for single items. Patient report measures showed some moderate correlations.

Conclusion

Existing SDM measures do not refer to a single construct. A gold standard is missing to decide whether any of these measures has the potential to indicate patient involvement.

Practice Implications

Pronounced heterogeneity of the underpinning constructs implies difficulties regarding the interpretation of existing evidence on the efficacy of SDM. Consideration of communication theory and basic definitions of SDM would recommend an inter-subjective focus of measurement.

Trial Registration

Controlled-Trials.com ISRCTN25267500.     

Evaluation of Construct and Criterion Validity for the ‘Liverpool Osteoarthritis in Dogs’ (LOAD) Clinical Metrology Instrument and Comparison to Two Other Instruments

Objective

To test the ‘Liverpool Osteoarthritis in Dogs’ (LOAD) questionnaire for construct and criterion validity, and to similarly test the Helsinki Chronic Pain Index (HCPI) and the Canine Brief Pain Inventory (CBPI).

Design

Prospective Study.

Animals

222 dogs with osteoarthritis.

Procedure

Osteoarthritis was diagnosed in a cohort of dogs on the basis of clinical history and orthopedic examination. Force-platform analysis was performed and a “symmetry index” for peak vertical force (PVF) was calculated. Owners completed LOAD, CBPI and HCPI instruments. As a test of construct validity, inter-instrument correlations were calculated. As a test of criterion validity, the correlations between instrument scores and PVF symmetry scores were calculated. Additionally, internal consistency of all instruments was calculated and compared to those previously reported. Factor analysis is reported for the first time for LOAD, and is compared to that previously reported for CBPI and HCPI.

Results

Significant moderate correlations were found between all instruments, implying construct validity for all instruments. Significant weak correlations were found between LOAD scores and PVF symmetry index, and between CBPI scores and PVF symmetry index.

Conclusion and Clinical Relevance

LOAD is an owner-completed clinical metrology instrument that can be recommended for the measurement of canine osteoarthritis. It is convenient to use, validated and, as demonstrated here for the first time, has a correlation with force-platform data.     

A systematic review of studies measuring and reporting hearing aid usage in older adults since 1999: a descriptive summary of measurement tools

Objective

A systematic review was conducted to identify and quality assess how studies published since 1999 have measured and reported the usage of hearing aids in older adults. The relationship between usage and other dimensions of hearing aid outcome, age and hearing loss are summarised.

Data sources

Articles were identified through systematic searches in PubMed/MEDLINE, The University of Nottingham Online Catalogue, Web of Science and through reference checking. Study eligibility criteria: (1) participants aged fifty years or over with sensori-neural hearing loss, (2) provision of an air conduction hearing aid, (3) inclusion of hearing aid usage measure(s) and (4) published between 1999 and 2011.

Results

Of the initial 1933 papers obtained from the searches, a total of 64 were found eligible for review and were quality assessed on six dimensions: study design, choice of outcome instruments, level of reporting (usage, age, and audiometry) and cross validation of usage measures. Five papers were rated as being of high quality (scoring 10–12), 35 papers were rated as being of moderate quality (scoring 7–9), 22 as low quality (scoring 4–6) and two as very low quality (scoring 0–2). Fifteen different methods were identified for assessing the usage of hearing aids.

Conclusions

Generally, the usage data reviewed was not well specified. There was a lack of consistency and robustness in the way that usage of hearing aids was assessed and categorised. There is a need for more standardised level of reporting of hearing aid usage data to further understand the relationship between usage and hearing aid outcomes.     

Agreement of Mammographic Measures of Volumetric Breast Density to MRI

Background

Clinical scores of mammographic breast density are highly subjective. Automated technologies for mammography exist to quantify breast density objectively, but the technique that most accurately measures the quantity of breast fibroglandular tissue is not known.

Purpose

To compare the agreement of three automated mammographic techniques for measuring volumetric breast density with a quantitative volumetric MRI-based technique in a screening population.

Materials and Methods

Women were selected from the UCSF Medical Center screening population that had received both a screening MRI and digital mammogram within one year of each other, had Breast Imaging Reporting and Data System (BI-RADS) assessments of normal or benign finding, and no history of breast cancer or surgery. Agreement was assessed of three mammographic techniques (Single-energy X-ray Absorptiometry [SXA], Quantra, and Volpara) with MRI for percent fibroglandular tissue volume, absolute fibroglandular tissue volume, and total breast volume.

Results

Among 99 women, the automated mammographic density techniques were correlated with MRI measures with R² values ranging from 0.40 (log fibroglandular volume) to 0.91 (total breast volume). Substantial agreement measured by kappa statistic was found between all percent fibroglandular tissue measures (0.72 to 0.63), but only moderate agreement for log fibroglandular volumes. The kappa statistics for all percent density measures were highest in the comparisons of the SXA and MRI results. The largest error source between MRI and the mammography techniques was found to be differences in measures of total breast volume.

Conclusion

Automated volumetric fibroglandular tissue measures from screening digital mammograms were in substantial agreement with MRI and if associated with breast cancer could be used in clinical practice to enhance risk assessment and prevention.     

A simple index for the high-citation tail of citation distribution to quantify research performance in countries and institutions

Background

Conventional scientometric predictors of research performance such as the number of papers, citations, and papers in the top 1% of highly cited papers cannot be validated in terms of the number of Nobel Prize achievements across countries and institutions. The purpose of this paper is to find a bibliometric indicator that correlates with the number of Nobel Prize achievements.

Methodology/Principal Findings

This study assumes that the high-citation tail of citation distribution holds most of the information about high scientific performance. Here I propose the x-index, which is calculated from the number of national articles in the top 1% and 0.1% of highly cited papers and has a subtractive term to discount highly cited papers that are not scientific breakthroughs. The x-index, the number of Nobel Prize achievements, and the number of national articles in Nature or Science are highly correlated. The high correlations among these independent parameters demonstrate that they are good measures of high scientific performance because scientific excellence is their only common characteristic. However, the x-index has superior features as compared to the other two parameters. Nobel Prize achievements are low frequency events and their number is an imprecise indicator, which in addition is zero in most institutions; the evaluation of research making use of the number of publications in prestigious journals is not advised.

Conclusion

The x-index is a simple and precise indicator for high research performance.     

Body Awareness: Construct and Self-Report Measures

Objectives

Heightened body awareness can be adaptive and maladaptive. Improving body awareness has been suggested as an approach for treating patients with conditions such as chronic pain, obesity and post-traumatic stress disorder. We assessed the psychometric quality of selected self-report measures and examined their items for underlying definitions of the construct.

Data sources

PubMed, PsychINFO, HaPI, Embase, Digital Dissertations Database.

Review methods

Abstracts were screened; potentially relevant instruments were obtained and systematically reviewed. Instruments were excluded if they exclusively measured anxiety, covered emotions without related physical sensations, used observer ratings only, or were unobtainable. We restricted our study to the proprioceptive and interoceptive channels of body awareness. The psychometric properties of each scale were rated using a structured evaluation according to the method of McDowell. Following a working definition of the multi-dimensional construct, an inter-disciplinary team systematically examined the items of existing body awareness instruments, identified the dimensions queried and used an iterative qualitative process to refine the dimensions of the construct.

Results

From 1,825 abstracts, 39 instruments were screened. 12 were included for psychometric evaluation. Only two were rated as high standard for reliability, four for validity. Four domains of body awareness with 11 sub-domains emerged. Neither a single nor a compilation of several instruments covered all dimensions. Key domains that might potentially differentiate adaptive and maladaptive aspects of body awareness were missing in the reviewed instruments.

Conclusion

Existing self-report instruments do not address important domains of the construct of body awareness, are unable to discern between adaptive and maladaptive aspects of body awareness, or exhibit other psychometric limitations. Restricting the construct to its proprio- and interoceptive channels, we explore the current understanding of the multi-dimensional construct and suggest next steps for further research.     

The Effect of Geographical Proximity on Scientific Cooperation among Chinese Cities from 1990 to 2010

Background

The relations between geographical proximity and spatial distance constitute a popular topic of concern. Thus, how geographical proximity affects scientific cooperation, and whether geographically proximate scientific cooperation activities in fact exhibit geographic scale features should be investigated.

Methodology

Selected statistics from the ISI database on cooperatively authored papers, the authors of which resided in 60 typical cites in China, and which were published in the years 1990, 1995, 2000, 2005, and 2010, were used to establish matrices of geographic distance and cooperation levels between cities. By constructing a distance-cooperation model, the degree of scientific cooperation based on spatial distance was calculated. The relationship between geographical proximity and scientific cooperation, as well as changes in that relationship, was explored using the fitting function.

Result

(1) Instead of declining, the role of geographical proximity in inter-city scientific cooperation has increased gradually but significantly with the popularization of telecommunication technologies; (2) the relationship between geographical proximity and scientific cooperation has not followed a perfect declining curve, and at certain spatial scales, the distance-decay regularity does not work; (3) the Chinese scientific cooperation network gathers around different regional center cities, showing a trend towards a regional network; within this cooperation network the amount of inter-city cooperation occurring at close range increased greatly.

Conclusion

The relationship between inter-city geographical distance and scientific cooperation has been enhanced and strengthened over time.     

Moving Away from Exhaustion: How Core Self-Evaluations Influence Academic Burnout

Background

Academic burnout refers to students who have low interest, lack of motivation, and tiredness in studying. Studies concerning how to prevent academic burnout are rare.

Objective

The present study aimed to investigate the impact of core self-evaluations on the academic burnout of university students, and mainly focused on the confirmation of the mediator role of life satisfaction.

Methods

A total of 470 university students accomplished the core self-evaluation scale, Satisfaction with Life, and academic burnout scale.

Results

Both core self-evaluations and life satisfaction were significantly correlated with academic burnout. Structural equation modeling indicated that life satisfaction partially mediated the relationship between core self-evaluations and academic burnout.

Conclusions

Core self-evaluations significantly influence academic burnout and are partially mediated by life satisfaction.     

Spatio-Temporal Changes of Lymphatic Contractility and Drainage Patterns following Lymphadenectomy in Mice

Objective

To investigate the redirection of lymphatic drainage post-lymphadenectomy using non-invasive near-infrared fluorescence (NIRF) imaging, and to subsequently assess impact on metastasis.

Background

Cancer-acquired lymphedema arises from dysfunctional fluid transport after lymphadenectomy performed for staging and to disrupt drainage pathways for regional control of disease. However, little is known about the normal regenerative processes of the lymphatics in response to lymphadenectomy and how these responses can be accelerated, delayed, or can impact metastasis.

Methods

Changes in lymphatic “pumping” function and drainage patterns were non-invasively and longitudinally imaged using NIRF lymphatic imaging after popliteal lymphadenectomy in mice. In a cohort of mice, B16F10 melanoma was inoculated on the dorsal aspect of the paw 27 days after lymphadenectomy to assess how drainage patterns affect metastasis.

Results

NIRF imaging demonstrates that, although lymphatic function and drainage patterns change significantly in early response to popliteal lymph node (PLN) removal in mice, these changes are transient and regress dramatically due to a high regenerative capacity of the lymphatics and co-opting of collateral lymphatic pathways around the site of obstruction. Metastases followed the pattern of collateral pathways and could be detected proximal to the site of lymphadenectomy.

Conclusions

Both lymphatic vessel regeneration and co-opting of contralateral vessels occur following lymphadenectomy, with contractile function restored within 13 days, providing a basis for preclinical and clinical investigations to hasten lymphatic repair and restore contractile lymphatic function after surgery to prevent cancer-acquired lymphedema. Patterns of cancer metastasis after lymphadenectomy were altered, consistent with patterns of re-directed lymphatic drainage.     

Effects of positive interactions, size symmetry of competition and abiotic stress on self-thinning in simulated plant populations

Background and Aims

Competition drives self-thinning (density-dependent mortality) in crowded plant populations. Facilitative interactions have been shown to affect many processes in plant populations and communities, but their effects on self-thinning trajectories have not been investigated.

Methods

Using an individual-based ‘zone-of-influence’ model, we studied the potential effects of the size symmetry of competition, abiotic stress and facilitation on self-thinning trajectories in plant monocultures. In the model, abiotic stress reduced the growth of all individuals and facilitation ameliorated the effects of stress on interacting individuals.

Key Results

Abiotic stress made the log biomass – log density relationship during self-thinning steeper, but this effect was reduced by positive interactions among individuals. Size-asymmetric competition also influenced the self-thinning slope.

Conclusions

Although competition drives self-thinning, its course can be affected by abiotic stress, facilitation and competitive symmetry.     

Mixed emotional experience is associated with and precedes improvements in psychological well-being

Background

The relationships between positive and negative emotional experience and physical and psychological well-being have been well-documented. The present study examines the prospective positive relationship between concurrent positive and negative emotional experience and psychological well-being in the context of psychotherapy.

Methods

47 adults undergoing psychotherapy completed measures of psychological well-being and wrote private narratives that were coded by trained raters for emotional content.

Results

The specific concurrent experience of happiness and sadness was associated with improvements in psychological well-being above and beyond the impact of the passage of time, personality traits, or the independent effects of happiness and sadness. Changes in mixed emotional experience preceded improvements in well-being.

Conclusions

Experiencing happiness alongside sadness in psychotherapy may be a harbinger of improvement in psychological well-being.     

Estimates of the Continuously Publishing Core in the Scientific Workforce

Background

The ability of a scientist to maintain a continuous stream of publication may be important, because research requires continuity of effort. However, there is no data on what proportion of scientists manages to publish each and every year over long periods of time.

Methodology/Principal Findings

Using the entire Scopus database, we estimated that there are 15,153,100 publishing scientists (distinct author identifiers) in the period 1996–2011. However, only 150,608 (<1%) of them have published something in each and every year in this 16-year period (uninterrupted, continuous presence [UCP] in the literature). This small core of scientists with UCP are far more cited than others, and they account for 41.7% of all papers in the same period and 87.1% of all papers with >1000 citations in the same period. Skipping even a single year substantially affected the average citation impact. We also studied the birth and death dynamics of membership in this influential UCP core, by imputing and estimating UCP-births and UCP-deaths. We estimated that 16,877 scientists would qualify for UCP-birth in 1997 (no publication in 1996, UCP in 1997–2012) and 9,673 scientists had their UCP-death in 2010. The relative representation of authors with UCP was enriched in Medical Research, in the academic sector and in Europe/North America, while the relative representation of authors without UCP was enriched in the Social Sciences and Humanities, in industry, and in other continents.

Conclusions

The proportion of the scientific workforce that maintains a continuous uninterrupted stream of publications each and every year over many years is very limited, but it accounts for the lion’s share of researchers with high citation impact. This finding may have implications for the structure, stability and vulnerability of the scientific workforce.