F . Forster et al .: Radioprotection 2024 , 59 ( 1 ), 55 – 64 57
2.5 . Synthesis of results
Based on the definition of the data items , comparable results of different articles were summarised by narrative synthesis . Items that were only assessed by one study are reported based on the results of the respective individual study .
3 Results
3.1 . Selection of sources of evidence
Altogether , 274 articles were identified . After removing one duplicate , 273 studies remained for title screening . 261 records were classified as non-eligible as they were identified as not relevant for our research question . Out of the twelve articles left , two did not address the population of interest , three did not focus on the required exposure and one did not include the outcomes of interest . Consequently , six studies were included in the full-text screening . The examination of the reference lists of these six studies led to three more records of interest . All nine sources met our inclusion criteria ( Fig . 1 ).
3.2 Characteristics of sources of evidence
All studies were conducted in Central / Western Europe . Five out of the nine records refer to a German study that was first published in 2009 with four more articles from the same study being published until 2015 ( Berg-Beckhoff et al ., 2009 , Berg-Beckhoff et al ., 2009 , Kowall et al ., 2010 , Berg- Beckhoff et al ., 2014 , Kowall et al ., 2015 ). The remaining studies were from Austria ( Leitgeb et al ., 2005 ), France ( Lambrozo et al ., 2013 ), the Netherlands ( Slottje et al ., 2017 ) and Switzerland ( Huss und Röösli , 2006 ). All studies had a cross-sectional design , included GPs and evaluated different outcomes in relation to our data items of interest . All studies used phone interviews or self-administered questionnaires . Additionally , the German study provided a long version and a short version of its questionnaire ( Tab . 1 ).
3.3 . Results of individual sources of evidence 3.3.1 . Level of information / knowledge
In the German study , GPs were asked about their EMFrelated knowledge regarding several EMF-related questions . Between those items , the percentage of physicians giving the correct answer varied between 26.4 % and 58.6 % of the participants . For every question , almost one third picked the option ‘ I don ’ t know ’. An additional latent class analysis revealed four kinds of answer types : GPs who answered correctly most of the time ; GPs who mainly chose ‘ I don ’ t know ’; GPs who were primarily able to answer questions on low-frequency EMF and GPs who answered health-related questions ( correct or incorrect ) and picked ‘ I don ’ t know ’ for the remaining questions ( Berg-Beckhoff et al ., 2009 , Berg- Beckhoff et al ., 2014 ).
A large part of the French GPs knew that mobile phone base stations , mobile phones , power lines and WiFi are the primary sources of EMF exposure in the general population . However , the authors concluded the GPs ’ knowledge concerning health risks in relation to EMF is limited
( Lambrozo et al ., 2013 ). The other three studies did not ask about GP ’ s knowledge levels .
3.3.2 . Attitude / perception of risk
In the Austrian study , a large part of GPs declared that in their opinion EMF can cause diseases . On a six-point Likert scale , 77 % chose a value on the ‘ agreement-half ’ and 33 % picked the highest value of agreement . The sources most mentioned as relevant for health-related effects of EMF were power lines , mobile phones , mobile phone base stations and wireless phones ( Leitgeb et al ., 2005 ). In the Swiss study , a smaller proportion , but still more than half of the GPs , estimated a relation between EMF and health-related symptoms to be plausible , while 29 % of them said it is not . Tinnitus , difficulties concentrating and neoplasms were the most stated possible health effects . Most relevant sources of EMF exposure mentioned by the GPs were mobile phones , wireless phones and other electric devices that are used close to the body ( Huss und Röösli , 2006 ).
Between 31.7 % ( short questionnaire ) and 57.5 % ( long questionnaire ) of the German GPs stated that in their view EMF exposure within legal exposure limits can cause health issues . The long questionnaire had an additional question on whether GPs believe that EMF can cause health issues . The difference to other questions was the phrasing ‘ do you think ...’ instead of ‘ are there ...’ and the missing of the additive ‘ within legal limits ’. Altogether , 54.3 % affirmed this statement . About half of the GPs ( 54.5 %) considered health issues related to EMF as psychosomatic . Most frequently stated symptoms were headache and sleep disorders . Most relevant EMF sources listed were mobile phones and mobile phone base stations ( Berg-Beckhoff et al ., 2009 , Berg-Beckhoff et al ., 2014 ).
The study from the Netherlands used a six-point Likert scale from fully disagree to fully agree for five different items . Almost two-thirds of the GPs disagreed that EMF exposure can cause health issues by choosing an answer on the ‘ disagreement-half ’. Furthermore , 73 % agreed to the statement that EMF-related health issues are psychosomatic most of the time . Regarding their last consultation , 18 % of GPs saw a plausible link between reported symptoms and EMF ( Slottje et al ., 2017 ).
3.3.3 . Relevance in practice and EMF-related consultations
Among Austrian GPs , 68 % reported that at least one of their patients has asked them about harmful effects of EMF . Moreover , 59 % indicated that they have had consultations with patients who even associated their health issues with EMF at least rarely or from time to time ( Leitgeb et al ., 2005 ). Similarly , in the Swiss study , 69 % of the GPs reported at least one consultation relating to EMF . In 77 % of these consultations , the suspicion of a link to EMF was stated by the patient . If advice was given , the GPs mainly recommended reducing exposure ( Huss und Röösli , 2006 ).
Those numbers were relatively similar to the German study , where 61.4 % of the GPs stated that health-related effects of EMF came up for discussion at least once . In almost three quarters of the consultations , patients themselves mentioned