Peer Review of “Medical Brain Drain from Southeastern Europe: Using Digital Demography to Forecast Health Worker Emigration”

<jats:p />


Specific Comments
1. Data

•
Have digital traces been used as an indicator of migration before?If so, please cite.
• I have concerns that some of the Google terms used indicate wider emigration rather than health worker migration.

•
If you want to refer to the wider emigration from Croatia to Germany/Austria (either via digital traces or via secondary data), you need to be clear that this is not health care worker (HW) migration you're referring to.I've compared HW migration with migration more generally in a recent paper [2].
• As an indication of health migration or HW migration intent, you should only focus on the doctor/nurse-specific information.The Introduction jumps around between source and destination countries.I would suggest that the paper discusses source countries and destination countries separately.
• Also, I think it's important to separate out push from pull factors and to be specific about the impact on each of the source countries; for instance, as an EU member state, Croatian doctors and nurses can freely migrate and work in Germany/Austria, but doctors/nurses from Serbia or B&H would need visas to work there, right?

•
The pandemic as a push factor is really interesting and an important issue to raise.
• Why use the term Western Balkan if it does not include Croatia?Better to use the countries that you're talking about (ie, Croatia, Bosnia and Herzegovinia, and Serbia).

XSL • FO
RenderX these issues are also relevant in European countries (see [3]).

•
In relation to Europe, the paper should connect back to the European Observatory books on HW migration from 2014/2015, which highlighted health worker migration from new EU member states to older EU member states [4].

•
In the Introduction, the paper should also connect with the wider literature on brain drain/health worker migration.

General Comments
Thank you for the revised manuscript; it is a much stronger paper, and the potential of digital demography in forecasting HW migration is now much clearer.• B&H is misspelled once in the Abstract.
• Throughout the paper, perhaps use a shortage of in place of a lack of.
3. Measuring health worker mobility is a challenge for most countries (not just Croatia, Bosnia, and Serbia), which is why registration and/or visa data from the destination country is often used as a measure of health worker emigration. 4. On page 4, you say that 65,288 nurses emigrated, and I think you're saying that there are more Croatian/Bosnian nurses in Germany than in Croatia and Bosnia; are you?Perhaps tighten up this sentence as it's a strong statement. 5. On page 7, the paragraph beginning in Austria is unclear.

3 .
Engaging with the wider literature • I think that the World Health Organization (WHO) Global Code on the International Recruitment of Health Personnel (2010), which is mentioned on page 11, should be more central to the paper, especially in relation to the concept of sustainability and the need for high-income countries to train and retain sufficient HWs to meet needs (article 5.4).• Perhaps the paper also needs to mention the WHO 2006 list or the WHO 2020 safeguard list, which lists countries with critical health care shortages.Your paper makes an interesting contribution in highlighting that JMIRx Med 2021 | vol. 2 | iss. 4 | e34078 | p. 1 https://med.jmirx.org/2021/4/e34078(page number not for citation purposes) Humphries JMIRx Med , perhaps mark the stock data versus the flow data (ie, differentiate between the overall number of nurses in Germany [stock] vs the number of nurses entering Germany [flow]). 7. On page 11, the section beginning what could the European Union do to address the problem, I think this should be moved out of the Introduction and into the Discussion. 8.I think you could bring one or two issues out in the Discussion that you've mentioned in the text already but could make more of: During the pandemic, the normal ways of data collection are simply too slow (particularly when EU countries are fast tracking health workers into the European Union).Your method is a really good way of generating timely insights into intent to migrate among health workers.As you mention in your paper, this should be useful for policy makers (but obviously only if they respond/react to the data).And then the next question for policy makers is how can they retain health workers during a pandemic?Increased salaries?Improved working conditions?Which links nicely to the section on what the European Union can do.
• I think that this method is a really interesting way of generating timely data on health worker migration.