A. Descriptive statistics:
- Demographics:
The survey collected data from 200 junior doctors divided into 181 doctors from 23 French-speaking countries and 19 from non-French-speaking countries.
The mean of age was 32,55 [min:23-max:56; STD=4,63]. Male/female ratio was 0,97. Distribution of participants by nationality showed a predominance of French 48 (24%), and Tunisian 48 (24%). The dominant country of work was France 28,5%.
Distribution according to academic degree for men was 18 (9%) interns, 18 (9%) residents, 4 (2%) Master students, 57 (28,5%) PHD and higher. For women, 39 (19,5%) interns, 53 (26,5%) residents, 10 (5%) master students, and 98 (49%) PHD and above. 111 (55,5%) participants had over 10 years of medical studies, while 84 (42%) studied from 5 to 10 years and only 5 (2,5%) less than 5 years.
Distribution of participants according to specialty and gender showed 133 (66,5%) neurologists with close male/female distribution respectively 63 (47,4%) and 70 (52,6%). Other specialties were represented by 67 (33,5%) participants with an equal male/female distribution 33 (49,2%).
2. Work and workplace characteristics:
- Financial income: 156 (78%) doctors had financial income. Distribution of income differed largely between countries of exercise. 44 (22%) participants were not paid and 71% among them worked in African countries.
- Human resources: The number of doctors with a university degree in departments was zero in 11% of cases, between 1 and 3 in 29% of cases, between 3 and 6 in 37% of cases, between 6 and 9 in 12% of cases, and above 9 in 11% of cases. Departments in which junior doctors worked had 70,6% professors, 66% specialists, 56,7% assistants and 44,3% associate professors. Teams were male predominant in 32,8%, female predominant in 32,3%, and an equal representation was observed in 28,3% of cases. Only 3(3,5%) of teams had a single gender. Teams with a single ethnic origin were most frequent (60%), compared to teams with two (17%) and three ethnicities (23%). Continental origins were African 66,5%, European 39,5%, Asian 11% and rarely American 4,5% and Australian 0,5%. 123 (61,5%) participants reported that they have equal scientific access regardless of their ethnicity. Meanwhile academic degree limited access to science for 129 (64,5%) junior doctors compared to their senior peers.
Workplace environment: Distribution of the workplace in the capital city (52%) and other cities (48%) did not differ significantly. 146 (73%) worked in university hospitals, 24 (12%) in specialized university hospitals, 7 (3,5%) in military hospitals, 15 (7,5%) in regional hospitals, 2 (1%) in primary care units and 6 (3%) in the private sector. Work-and-life balance: 70,7% worked between 6 to 10 hours, while 7,6% worked less than 6 hours and 21,7% worked over 10 hours. The number of night-shifts each month was below 4 in 40,2%, between 4 and 8 in 33,2%, and over 8 in 2%. Introspection about the importance of diversity in the scientific access: 33% participants valued diversity, while 15,5% did not consider it important in the workplace. 51,5% participants didn’t know if diversity had a positive impact on scientific access. French participants disagreed 11 (23%) or were not sure 30 (63%), while Tunisian participants agreed 15 (31,2%) or were not sure 29 (60%) about the importance of diversity. Gender introspection on diversity showed that male participants approved (38%) or doubted (42%), while female participants often doubted (60%) the importance of diversity. Scientific access for education and publishing: The main scientific resources used for medical education were Journals 82,9%, Congresses 79,4%, Sci-Hub 74,5% followed by individual learning 63,3%, Books 60,8%, CME 47,7%, non-formal resources 18,1%, and exchange summer/spring schools 14,1%. 119 (59,5%) defended their thesis. Within this group, 81 (68,1%) reported difficulties in publishing their thesis in journals. 103 (51,5%) wrote at least one article, while 23 (11,5%) had one ongoing, and 73 (36,5%) did not write any. The number of articles written was in 23,5% of cases between 2 and 5, and in 13% of cases only one. 9% wrote between 6 to 10 articles, 5,5% wrote between 11 to 20, and 3% wrote over 20 articles. Distribution of published articles according to gender showed that more than a half of women (55%) published between 2-5 articles whereas, most of people who published more than 10 articles were men (88%). Working in France and Tunisia was linked to a higher number of published articles within the range of 1 to 5. 53% of junior doctors in Madagascar reported a number of published articles above 6. Academic grade showed a general predominance in the number of articles published below 5, and a greater number of publications after PhD. These articles were published in a journal with an impact factor below 1 in 32% and between 1 and 4 in 14,5%. The main journal selection criteria were a high-impact factor (67,4%), a journal in the same specialty (57%), free submission (55,4%), similar research theme (42,5%), short reviewing time (32,6%) and reviewing quality (29%). The minimum reviewing time for an article is divided equally between the periods 1 to 4 months (39%) and 5 to 8 months (38%). Difficulties in writing a scientific paper were divided into 4 categories: linguistic (56,5%), financial (64,7%), scientific (55,3%), and logistic (65,3%). Writing in a foreign language 93(70,5%) was the main difficulty reported, followed by scientific writing 75(56, 8%), scientific editing 65(49,2%) and choice of pertinent information 55(41,7%). Financial difficulties included expensive submission fees 106(79, 1%), the absence of university financial support 99(73,9%), expensive articles 92(68,7%) and books 73(54,5%), the absence of financial support modalities of submission for low- and middle-income countries 60(44,8%), and currency exchange restrictions 35(26,1%). Financial difficulties in Madagascar limited scientific publishing (P=0,007, OR 10,04[13 ;77,7]) in opposition to France (P=0,000, OR 0,084[0,35; 0,2]). The percentage of scientific difficulties was over 50% in all categories and was a result of difficulties using softwares like Excel, SPSS, Endnote and Zotero 71,4%, the slow update of national scientific resources 63,7%, and the uncertainty about the validity of scientific information 59,1%. Logistic difficulties included different submission criteria in each journal 61,1%, complicated submission procedures 57,9%, slow reviewing 50,5%, and difficulties in selecting the appropriate journal 45,8%. Motivation and availability to write accounted for 32,1% and 33,7% of cases while use of internet resources, and relational difficulties due to ethnicity gender and academic grade were the least reported.
B. Analytical statistics:
Access to science for medical training:
The association between geographic, academic factors and educational resources used in science is detailed in Table 1.
Access to science for publishing:
- Impact of demographics on scientific publications:
Impact of age: The mean of age was close in the different groups and didn’t seem to have an impact on scientific publications (articles and thesis) and difficulties. The impact of age on the scientific publishing and scientific education, is represented in figure 1.
- Impact of gender and ethnicity:
We analyzed the differences in scientific publishing for both genders in France and Tunisia. We noticed that frequencies of submitted thesis were close in both countries for both genders (20 to 26% for men and 30 to 33% for women). The percentage of published articles was, however, more important in the man's group in Tunisia 80% compared to the woman's group 48%. This difference was not identified for doctors working in France (31% for men and 33% for women). The number of published articles didn’t surpass 2 to 5 for both genders in Tunisia (50%-69%). Meanwhile, in France, the majority published between 2 to 5 (47%-54%) a group of participants could publish between 6 to 10 articles (15%-20%).
Scientific publishing according to monthly income, academic grade, ethnical origins and gender is represented in figure 2.
- Impact of specialty and number of years of education on scientific publishing:
The maximum number of articles published for junior neurologists and doctors of other specialties was between 2 and 5, with respective percentages of 46% and 35%. The number of published articles increased according to the number of years of education. I was more difficult to publish a thesis in the field of neurology 54(66,7%) compared to other specialties 27(33,3%). Difficulties increased in parallel to the number of years of education. In fact, 78 % of participants who studied for more than 10 years reported having difficulties publishing their articles compared to only 2.5% of those with 2 years of education. Income did not seem to influence these difficulties that were present in highly paid and non-paid groups.
- Workplace environment impact on publishing; including the number of senior doctors holding and academic degree, gender, ethnic distribution of senior doctors and workplace location: Thesis publication difficulties were more important in teams that had less than 6 medical university hospital teachers and gender did not influence these difficulties. The number of articles published depended on the workplace location and the health structure type. The role of workplace environment in publishing articles is represented in figure 3.
Work-and-life balance affected positively thesis and articles publication in groups working between 6 to 10 hours a day and having less than 4 on-call shifts per month (figure 4).
Influence of journal selection criteria on publications: For participants who published several articles between 2 and 5, choosing a high or low-impact factor journal, a paying journal, or reviewing duration did not change chances of manuscript acceptance. However, choosing a non-paying journal was correlated with 32,7% chances of having published manuscripts. Also, reviewing quality, same specialty, same research theme, known editorial board members and special offers increased chances to publish by 25% to 36,5%. Submitting manuscripts in journals using one’s native language increased chances of publishing to 50%.
- Analysis of difficulties associated with scientific publishing by nature, country and academic degree (table 2):