Google Health Search Trial (GHST)

• ClinicalTrials.gov Identifier: NCT03073746
• Recruitment Status: Completed
• First Posted: March 8, 2017
• Last Update Posted: April 16, 2019
• Sponsor: Johns Hopkins University
• Collaborator: Google LLC.
• Information provided by (Responsible Party): Johns Hopkins University

Study Description

Brief Summary:

There are large differences in knowledge between patients and healthcare providers (i.e. physicians, physician assistants and nurse practitioners), and there is a strong interest on the part of both industry and academia to reduce the gap in knowledge between patients and healthcare providers. Currently, about 1 in 20 searches on Google are health related. Among internet users, 72% reported searching for health information, and among persons who use mobile phones, 31% of cell phone users and 52% of smartphone users have looked up health or medical information. Oftentimes, patients will search on Google or other search engines in order to find health conditions that explain their symptoms prior to visiting their healthcare provider. With the launch of Google's new health search tools for mobile devices (i.e. smartphones, tablets, etc.) it is important to understand how patients use these search platforms and what their effects are on clinical encounters. The main objective of this study is to understand the accuracy of differential diagnoses generated by Google searching; the investigators hypothesize that searching on Google using a tablet or mobile device will be more accurate than not using any search tool, and that the new health search experience will improve accuracy over the standard search platform.

Condition or disease	Intervention/treatment	Phase
Diagnoses Disease	Other: Health Search; Other: Standard Search	Not Applicable

Study Design

• Study Type: Interventional (Clinical Trial)
• Actual Enrollment: 300 participants
• Allocation: Randomized
• Intervention Model: Parallel Assignment
• Intervention Model Description: Participants presenting with new symptoms to either an urgent care center or emergency department will be randomized in a 1:1:1 ratio to one of three groups. 1. Access to Health Search (i.e. Google's new health search platform for mobile devices; Health Knowledge Panels and Symptom Search Tool); 2. Access to Standard Search (i.e. prior version of Google search); 3. Usual Care. Groups 1 and 2 will use a tablet or mobile phone to search while group 3 will not.
• Masking: Single (Care Provider)
• Masking Description: Healthcare providers will not have knowledge of the interventions assigned to individual participants.
• Primary Purpose: Other
• Official Title: Accuracy of Differential Diagnosis From Google Health Search and Impact on Clinical Encounters
• Actual Study Start Date: September 7, 2016
• Actual Primary Completion Date: August 9, 2017
• Actual Study Completion Date: August 15, 2017

Arms and Interventions

Arm	Intervention/treatment
Experimental: Health Search; Access to Health Knowledge Panels and Symptom Search Tool.	Other: Health Search
Experimental: Standard Search; No access to Health Knowledge Panels and Symptom Search Tool, but access to prior version of Google search.	Other: Standard Search
No Intervention: No Search; No access to Health Knowledge Panels, Symptom Search Tool, prior version of Google search, or mobile device.

Outcome Measures

Primary Outcome Measures :

1. Accuracy of patient's pre-visit differential diagnosis after Google searching on a tablet or mobile phone. [ Time Frame: Assessing diagnostic accuracy from baseline to up to one hour post-clinical encounter. ]
   Proportion of patient's pre-visit differential diagnosis meeting the criteria of "match" with clinician post-visit differential diagnosis (i.e. Two of the three conditions on the clinician's differential diagnosis were more likely to be present in the patient's pre-visit assessment when a tablet or mobile phone was present).

Secondary Outcome Measures :

1. Accuracy of differential diagnosis of Health Search vs. Standard Search [ Time Frame: Assessing diagnostic accuracy from baseline to up to one hour post-clinical encounter. ]
   Proportion of patient's pre-visit differential diagnosis meeting the criteria of "match" with clinician post-visit differential diagnosis comparing Arm 1 (i.e. Health Search) vs. Arm 2 (i.e. Standard Search).

Other Outcome Measures:

1. Changes in anxiety on a visual analog scale [ Time Frame: Assessing a change in anxiety from baseline to up to one hour post clinical encounter. ]
   Comparing changes in anxiety (pre-clinical encounter vs. post-clinical encounter) among treatment arms. (Health Search vs. Standard Search).

Eligibility Criteria

• Ages Eligible for Study: 18 Years and older (Adult, Older Adult)
• Sexes Eligible for Study: All
• Accepts Healthy Volunteers: No

Criteria

Inclusion Criteria:

• Adult patient >= 18 years of age
• Presenting for care at urgent care or emergency department
• In search of a diagnosis to explain a new symptom or group of symptoms

Exclusion Criteria:

• Not literate, defined as lacking the ability to read and write (self-reported)
• Non-English Speaking (self-reported)
• Not mentally competent to provide consent due to inability to understand relevant information due to deficit in intelligence (e.g. mental retardation), memory (e.g., advanced dementia or significant delirium), or attention span (e.g., Attention Deficit Disorder (ADD) or mania) based on prior documentation in medical records or as judged by the researchers)
• Unable to use phone/tablet for any mental or physical impairment (e.g., blind) (self-reported)
• Exacerbation of chronic condition (self-reported)

Contacts and Locations

Locations

United States, Maryland

Johns Hopkins Bayview Medical Center

Baltimore, Maryland, United States, 21224

Johns Hopkins Hospital

Baltimore, Maryland, United States, 21287

Sponsors and Collaborators

Johns Hopkins University

Google LLC.

Investigators

• Principal Investigator: Seth S Martin, MD, MHS; Johns Hopkins School of Medicine

More Information

Publications:

"Health Fact Sheet." Pew Research Center, Washington D.C. (December 16, 2013). http://www.pewinternet.org/fact-sheets/health-fact-sheet/.

• Responsible Party: Johns Hopkins University
• ClinicalTrials.gov Identifier: NCT03073746
• Other Study ID Numbers: IRB00088552
• First Posted: March 8, 2017
• Last Update Posted: April 16, 2019
• Last Verified: April 2019

• Studies a U.S. FDA-regulated Drug Product: No
• Studies a U.S. FDA-regulated Device Product: No

Keywords provided by Johns Hopkins University:

Google; Knowledge Panels; Health Search; Online; Internet; Smartphone; Tablet; Digital; Mobile; Symptoms; Differential diagnosis; Diagnosis; Anxiety

Additional relevant MeSH terms:

Disease; Pathologic Processes