For Authors

Content and Audience

AI and Healthcare is a multidisciplinary, peer-reviewed journal dedicated to the exploration and acceleration of analysis and discussion on the improvement of health and health care through the application of artificial intelligence technology.

We encourage private and public researchers, decisionmakers, policymakers, and innovators to contribute papers. We wish to incorporate diverse viewpoints and perspectives in technology, government, care delivery, care management, and, importantly, patient experience.

General Contents

AI and Healthcare journal authors should direct their papers to a diverse national audience interested in present and future issues related to the development and application of artificial intelligence in healthcare.

New data, methodologies, research findings, reports of major new trends, examples of adoption and diffusion, case studies, perspectives, and proposals for reform are encouraged.

AI and Healthcare values discussion on both the development of data, technology, and the effect on health and healthcare. Thus, we encourage authors to discuss both the advancement of AI and the impact on healthcare. Whenever appropriate, data and methods will be assessed in the peer review process. We encourage authors to provide more methodological detail in the form of an appendix or other explanatory section for review purposes only. We also encourage authors to discuss the health and healthcare implications of technology, as well as guidance to further research, adoption, and diffusion.

Manuscript Types

Several types of manuscripts may be submitted to AI and Healthcare, as described below. If you have any questions about which option to choose, please contact the editorial office at hello@aiandhealthcare.org.

I. Research Article: 2,000 to 5,000 words, up to 4 exhibits (tables and figures). Research articles can be:

  • Based on original/primary research or secondary research
  • Descriptive or analytical
  • Quantitative, qualitative, or mixed methods
  • Case studies

Research articles proceed according to a standard formula, including sections presenting an introduction, data/methods, results/findings, discussion/conclusions.

II. Analysis: 3,000 to 5,000 words, up to 4 exhibits (tables and figures)

Analysis papers draw on known facts/data and existing research relevant to a particular AI technology or implementation question with the aim of answering or promoting understanding of the question. These papers can include but are not limited to formal policy analysis. Analysis papers are expected to draw conclusions based on clear reasoning from known evidence, reflect a complete and current understanding of that evidence, and be responsive to counter arguments. This category covers systematic and literature reviews and meta-analyses.

III. Data: 2000 words or less, up to 6 exhibits

Data articles are short papers that highlight data that “speak for themselves” relative to important policy issues or topics. They should shed light on some important question and be “worth knowing.” They do not typically test hypotheses, rely on sophisticated statistical methods, or include lengthy policy discussions. We aim to present new data or new analyses of existing data that are reliable and credible and that promote understanding among nonexperts on important, policy-relevant topics. We encourage work based on underused or new data sources. More information is in our announcement for the series.

IV. Commentary, Overview, Policy Insight, Perspective: 2,000 to 7,000 words, up to 4 exhibits

Commentaries address important, current challenges in artificial intelligence and healthcare. They present a “point of view.” Commentaries are expected to build a thesis or argument through analysis of established facts that are relevant to the problems that they are addressing. While Commentaries are not expected to present systematic reviews of existing evidence/literature, they should present a responsible and balanced treatment of the state of knowledge on the topic.

Manuscript Style

General: Authors should refer to the Chicago Manual of Style (University of Chicago Press) as a style guide for the text portion of their paper. 

AI and Healthcare uses the Vancouver style (following the Uniform Requirements for Manuscripts Submitted to Biomedical Journals) for references. Endnotes, including both sources and explanatory matter, should be kept to a minimum and numbered in the order in which they appear in the text, not alphabetically.

Sources: 
Journal-article citations must include the volume, issue number, and year of publication as well as the article page numbers. Book and report citations must include the pertinent details of publication (publisher, location, year). References to online documents must contain complete details (author’s name if relevant, name of document, and citation information), not simply a URL. Improperly formatted notes may be returned to the author for revision, possibly delaying a paper’s publication. All sources must be provided in English. Authors are responsible for the accurate translation of foreign-language sources’ publication details into English during the revision process.

Endnotes: 
Here are some of the highlights of the Vancouver endnote style:

  • AI and Healthcare does not use footnotes; all notations and references should be integrated into one section of consecutively numbered endnotes that are listed in the order in which they are called out in the text.
  • Use Arabic, not Roman, numbering.
  • References are individually assigned a unique “callout” number, and note numbers are repeated in the text whenever the source material applies.
  • Sentences may contain as many endnotes as are relevant, in whatever position they are most directly pertinent to the text where they apply.
  • If a publication or article has up to six authors or editors, all should be listed; if a publication or article has more than six authors, the first six are listed, followed by “et al.”

Many authors use software packages to prepare their lists of references. The three major packages (EndNote, ProCite, and Reference Manager) all contain an output style or bibliographic style for Vancouver. Authors maintain the responsibility of using the program(s) appropriately

Endnotes should conform to the following examples.

1. Sparer MS, Brown LD. Uneasy alliances: managed care plans formed by safety-net providers. Health Aff (Millwood). 2000;19(4):23-35.

2. Cowan CA, Lazenby HC, Martin AB, McDonnell PA, Sensenig AL, Smith CE et al. National health expenditures, 1999. Health Care Financ Rev. 2001 Summer; 22(4):77-110.

3. Starr P. The social transformation of American medicine. New York: Basic Books; 1982.

4. Harris G. FDA orders recall of intravenous pumps. New York Times (Washington Final). 2005 Jun 22;Sect. A:12 (col. 1).

5. Corrigan JM, Kohn LT, Donaldson MS, editors. To err is human: building a safer health system. Washington: National Academies Press; 1999.

6. Anton B, Nelson R. Literacy, consumer informatics, and health care outcomes: interrelations and implications. In: Park HA, Murray P, Delaney C, editors. Consumer-centered computer-supported care for healthy people. Amsterdam: IOS Press; 2006. p. 49-53.

7. Patrias K. Computer-compatible writing and editing. Paper presented at: Interacting with the digital environment: modern scientific publishing. 46th Annual Meeting of the Council of Science Editors; 2003 May 3-6; Pittsburgh, PA.

8. Collins SR, Kriss JL, Davis K, Doty MM, Holmgren AL. Squeezed: why rising exposure to health care costs threatens the health and financial well-being of American families [Internet]. New York: Commonwealth Fund; 2006 Sep [cited 2006 Nov 2]. 34 p. Available from: 

http://www.cmwf.org/usr_doc/Collins_squeezedrisinghltcarecosts_953.pdf.

More extensive information on citing in Vancouver style is available online. We recommend the following sources:

National Library of Medicine, “International Committee of Medical Journal Editors Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Sample References,” http://www.nlm.nih.gov/bsd/uniform_requirements.html

National Library of Medicine, “Citing Medicine,” http://www.ncbi.nlm.nih.gov/books/bv.fcgi?rid=citmed.TOC&depth=2

Monash University, “Vancouver style (uniform requirements for manuscripts submitted to biomedical journals),” http://www.lib.monash.edu.au/tutorials/citing/vancouver.html

For specific questions on endnote style, please email: hello@aiandhealthcare.org

Exhibits: Here are some key style notes for article tables and figures:

  • Call out In text: Exhibits should be mentioned (“called out”) in the flow of the text, not simply as an added note suggesting their placement.
  • List Of Exhibits: There should be a comprehensive list of exhibits at the end of the text document, with each item in the list indicating whether the exhibit is a table or a figure, the title of the exhibit, and the source(s) for the data/information presented.
  • Sources: Sources must be provided, even if the source is the authors’ analysis. If this is the case, please also specify what data the authors are analyzing (for example, “Authors’ analysis of data from the Medical Expenditure Panel Survey, 2014”).
  • Labeling: Exhibits should have clear titles (“captions”), subheading labels, and (for figures) legends, and they should be understandable at a glance.
  • Simple Presentation: Overly complicated exhibits or exhibits that are not formatted for nonexperts — for example, raw regression output — may be removed before publication and placed in an appendix for online viewing.
  • Consecutive Numbering: Exhibits must be numbered consecutively as they appear, not separately based on whether they are a table or a figure.
  • Layout: Tables should not be excessively horizontal in layout.
  • Permissions: If the work in the exhibit is not original and has been published elsewhere, the author is expected to obtain copyright permission from the publisher and is responsible for any fees that may be required.