Objectives: To determine whether it is possible to contact authors of previously published papers via email.
Methods: A cross sectional study of the Emergency Medicine Journal for 2001.
Results: 118 articles were included in the study. The response rate from those with valid email addresses was 73%. There was no statistical difference between the type of email address used and the address being invalid (p=0.392) or between the type of article and the likelihood of a reply (p=0.197). More responses were obtained from work addresses when compared with Hotmail addresses (86% v 57%, p=0.02).
Conclusions: Email is a valid means of contacting authors of previously published articles, particularly within the emergency medicine specialty. A work based email address may be a more valid means of contact than a Hotmail address.
Statistics from Altmetric.com
The internet and email are now an integral part of most people’s lives. A survey of 771 doctors in the West Midlands in 2001 showed that 65% of respondents used email, 92% for communicating with friends and family and 61% for communicating with work colleagues.1 They also showed that email use was highest in the 20–29 year age group and lowest in those aged over 60 years.1
Within the emergency medicine community, Alves and Szucs surveyed 1752 physicians from the 1999 American College of Emergency Physicians Membership Guide. Seventy nine per cent had valid email addresses but only 42% responded to their email. They showed a 13.6% improvement in response with individualised compared with bulk mail.2
Most papers currently published in the Emergency Medicine Journal list an email address as a means of contacting the principal author. However, to the author’s knowledge, there has been no research into the use of email in this way.
The aim of this study was to determine whether it is possible to contact authors of previously published papers via email. In addition, to determine the type of email address used, the ability to deliver email to the author’s address, and the response rate of authors when compared with the type of paper published and the type of email address used.
The principal author’s email contact details were obtained from all original, review, and best practice articles and case reports published in the Emergency Medicine Journal in 2001. The review and best practice articles were categorised together as they were thought to be of a similar nature.
Emails were sent to the addresses specified in each of the papers from an address (yahoo.com.au), which was established specifically for this study. These individualised emails asked, in most cases, a short question regarding the authorship of the study or for a reply, where the former was thought to be inappropriate.
The number of emails returned undeliverable and the number of replies were noted. A follow up email was sent at two weeks to initial non-responders.
In addition to descriptive statistics, the χ2 test was used to examine the association between categorical variables. Although the significance level was set at 0.05, the Bonferroni method was used to adjust for individual α levels when multiple testings were performed.
Altogether 118 articles were included in the study; 8.5% were best practice or review articles, 51.7% were original articles, and 39.8% were case reports.
Of the 118 articles, 111 (94%) provided an email address for contacting the principal author; 31.5% of these were work addresses, 45% home, and 23.5% Hotmail.
Fifteen of the emails were undeliverable, indicating that 86.5% of the email addresses were valid. There was no statistical difference between the type of email address used and the address being invalid (p=0.392).
A total of 70 authors responded, 63% of all authors providing an email address and 73% of those with valid email addresses. There was no statistical difference between the type of article and the likelihood of a reply (p=0.197). Figure 1 shows the numbers of articles and the number of replies received for each type of article.
On comparing the likelihood of a reply with the type of email address the results were just insignificant (p=0.065). Sub-analysis showed that more responses were obtained from work rather than Hotmail addresses (86% v 57%, p=0.02). This result is marginally insignificant, as the new α level would be 0.0167 after the Bonferroni adjustment. Figure 2 shows the number of articles, the number of valid emails, and the number of replies for each type of email address.
A response rate of 73% is better than previous studies have shown when the contact has been totally unsolicited (42%–58%).2,4 However, it compares favourably with previous results where contact has been partly encouraged.5 The authors of papers in the Emergency Medicine Journal may expect to be contacted by email by people interested in their paper, and so may be more likely to respond. The use of an individualised email rather than bulk mailing may have also contributed.
We can assume that email will be an increasingly important means of communicating with our professional colleagues as younger doctors begin to practice,2 and this may lead to further increases in response rates.
This study has also shown that there may be a more significant chance of reply from a work email address rather than a Hotmail address. This could be because web based email accounts receive more unsolicited email (“spam”) than work accounts, so recipients may be more likely to simply delete email from unfamiliar people. It could also be because work based accounts are accessed more frequently than web based accounts.
There was no significant difference between those emails undeliverable and the type of email address. This may have been attributable to the small number of papers involved and the study not having the power to detect any significance. The emails were returned for a variety of reasons; the most prevalent reason was an inactive or invalid account, presumably because of infrequent use by the user or because of the author moving positions or regions and losing their work based address.
This study is limited to a specific population and the results may not be valid for other health professionals publishing in other journals. The use of a larger sample size would increase the power of the study and enable the question of work based email addresses being more valid to be answered more definitively. A web based email account was used for convenience and this may have contributed to a lower response rate as recipients may have thought the sender was sending them unsolicited “spam”.
In conclusion, email is a valid means of contacting authors of previously published articles, particularly within the emergency medicine specialty. Work based email addresses may be a more valid means of contact than a Hotmail address. Further research is required to establish whether email can replace traditional postal contact, when attempting to communicate with our professional colleagues.
The author would like to thank Sing Kai Lo, PhD, statistician from the Institute of International Health, University of Sydney for his help in analysing the data and preparing the statistical analysis.
The author is the sole contributor to this paper and will act as guarantor
Conflicts of interest: none declared.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.