Author’s response to reviews Title: Physicians' perceptions on the impact of telemedicine on recruitment and retention in underserved areas: a descriptive study in Senegal Authors:

The bold conclusion they have drawn from this paper is not right and not according to the title and results. "All the physicians surveyed agreed, but with different emphasis, that telemedicine is not the only factor that determines recruitment and retention...." I didn't read anything about perception in this conclusion: We changed the conclusion. Introduction: the introduction section is very shallow and it is not based on sound arguments. This is one of the main flaws of this paper. This part should be developed with more arguments. How telemedicine could influence physicians’ intent to stay working in underserved areas. How telemedicine would improve motivation of physicians, facilitate professional integration, improves knowledge management, access to specialist knowledge and improvement of consultations as well as medical care efficiency needs to be well argued. Authors should also provide a better argument what this study will provide to the existing knowledge base, and find a better justification of why this study is important: We introduced new sentences and made other changes. "Several studies have demonstrated its positive impact on physicians' recruitment and retention (4, 5). Gagnon and colleagues reported positive findings in nine of the 13 studies included in their literature review (6). Dolea and colleagues found similar results (7)" What are these similar results? It is not clear: We deleted this sentence. "This study explores their perceptions and the factors influencing their perception." This is not clear: Why is it unclair? This comment is not very helpful for us. We however made some changes. Methods: The authors mentioned "....random sampling to purposive select...." This is a qualitative study, how can you combine random sampling with purposive sampling? Radom sampling is a probability sampling method while purposive is a non-probability sampling method: We deleted purposive from the sentence. In addition, the authors have selected 30 physicians in each level of facility (Public hospitals and district health centres) while these two facilities have different number of physicians. How do you fix this number to 30 each? Why not 5, 10, or 20 or less or more? Normally, in qualitative studies, you interview potential respondents until you reach the level of "saturation". Besides that, the numbers are not based on proportional allocation. This is also reflected in the results section: This study was combined with another study that focused on physicians’ beliefs. For this second study, we used the theory of planned behavior. This theory recommends to include 30 participants when studying individuals’ beliefs. That’s why we selected 30 physicians by group. The authors mentioned they use "interview schedule" as a data collection tool. What is interview schedule? We mean interview guide. We changed schedule by guide. The interview questions displayed in table 1 are superficial and are not arranged according to level of difficulty. Ordering of the questions should have been from general questions (Q#4) to specific questions as in number 1 and 2 and 3. Q#1 and Q#4 are leading questions: Why should we arrange our questions according to level of difficulty when we need specific answers? Our Q#2 is related to Q#1 and Q#4 to Q#3. Q#2 have to follow Q#1 as Q#4 have to follow Q#3. It doesn’t make sense to ask Q#2 before Q#1 and Q#4 before Q#3. Concerning Q#5, it should come after Q#3 et Q#4. The Duplantie and colleagues framework should have been clearly described in the methods section: It is not a framework. It is a classification and the different classes are mentioned in the text. See the reference. Results: The arbitrary selection of physicians in the sample is reflected in Table 2 of the results section. There are no GPs working in public hospitals represented in the sample. The selection of participants is not systematic; it is rather haphazard: Most of the physicians working in public hospitals are specialist physicians. GPs are really rare in Senegal public hospitals. It is easier to meet trainee physicians than GPs in Senegal public hospitals. Another major problem is that the physicians working outside Dakar are underrepresented in the sample, while the study is about recruitment and intention to stay in underserved areas. A physician working in Dakar would describe the situation as a physician working outside Dakar: That is not really true. The physicians working in district health centres are more numerous outside-Dakar than in Dakar (see Table 2). The reason is simple. Most of the district health centres are outside-Dakar. Most of the physicians working in public hospitals work in Dakar, because most of the public hospitals are in Dakar. Table 2 is not linked with the text: Please see the sentence before the table 2. Positive Perceptions: please revise the statement from line 36 to 46, page 9 "Finally........experts". It is very confusing: we did it. The paper has a mix-up of eLearning with telemedicine. The paper is about elearning not about telemedicine. This is reflected in the results for example: "Of course, it could retain us because we don't need to travel to Dakar to learn" (Male specialist physician: Don’t confuse teletraining and elearning. Telemedicine include teletraining. Here, we are talking about teletraining which involves at least an expert (usually from Dakar) and a physician (from Outside-Dakar). Overall the results are superficial as they are based on a superficial interview questions: Our questions are may be superficial, but our results are very important for Senegal Health System which felt to equitably distribute its physicians. Our results can guide telemedicine strategies in Senegal. For example, they tell Senegal’s health authorities to not use telemedicine alone and to combine it with other interventions to better recruit and retain physicians in underserved areas. This is very important for us. It is also important to know that 60% of our respondents thought that telemedicine could have an impact on their recruitment and retention. The reasons why it could have an impact on their recruitment and retention are equally known. Discussion. Overall the discussion is better than the other subtopics and it discussed the results very well: Thanks. Line 2931, page 17 "They coincide with the results of the authors who stated that isolation, fatigue and stress at work can prevent physicians' retention (5)." This is not right: I deleted this sentence. Also rephrase the next statement from line 31 to 37: It was rephrased. Conclusion: The conclusion is nothing new. It is very superficial and there is nothing surprising new finding: Our conclusion is maybe nothing new, but it is adapted to our study, which shows that the majority of the interviewed physicians had a positive perception about the impact of telemedicine on their recruitment and retention in underserved areas. This is very important because we have here the first information on the perceptions of Senegal’s physicians about the impact of telemedicine on their recruitment and retention in underserved areas. We have also the information that telemedicine is not the only factor that determines the decision of Senegal’s physicians to work and stay in underserved areas. Finally, we have the first list of the other factors that determine the decision of these physicians to work and stay in underserved. We suggest that telemedicine should be used with caution as a means to recruit and retain physicians in underserved areas and we propose to combine it with other recruitment and retention strategies. This study, its results and its conclusions are very important for us, Senegal Health System and Senegalese people who are suffering from the misdistribution of physicians and the lack of adapted telemedicine system. Everyone in the field would reach the same conclusion without doing this study. The authors should also provide a more substantial outlook of their study results: Maybe, but it is better to study and our study is providing the first information on the perceptions of Senegalese physicians on the impact of telemedicine on their recruitment and retention in underserved areas. Nobody focused on these questions before. We modified our conclusion. Declarations: List of abbreviations line 9 is in French, please change that to English if you are publishing your article in English: I changed it. If you also have only 2 abbreviations, there is no need to have a list of abbreviation section: I deleted the list of abbreviations. Authors' contributions needs to be revised. Please see the Vancouver authorship rule to write authors contribution. The co-authors (who are the co-authors). Mention the name of authors with initials: I mentioned the name of authors with their initials specifying the lead author (Birama Apho Ly) and co-authors (Ronald Labonté, Ivy Lynn Bourgeault and Mbayang Ndiaye Niang). The paper needs to be proofread by a native English speaker for inappropriate use of words, outstanding grammatical and punctuation errors: We forgot two French words in the text, but two of us (Ronald Labonté and Ivy Lynn Bourgeault) are native English speakers. Five other English speakers revised this paper during my thesis defense last year at the University of Ottawa. Additionally, an editor revised it.