Confidential: For Review O
nly
The Effectiveness of Symptom Checkers for Self-Diagnosis
and Triage: Beyond “Googling” Symptoms
Journal: BMJ
Manuscript ID: BMJ.2015.025489.R1
Article Type: Research
BMJ Journal: BMJ
Date Submitted by the Author: 14-May-2015
Complete List of Authors: Semigran, Hannah; Harvard Medical School, Health Care Policy Linder, Jeffrey; Brigham and Women's Hospital, Medicine; Harvard Medical School, Gidengil, Courtney; RAND Corporation, ; Boston Children's Hospital, Infectious Diseases
Mehrotra, Ateev; Harvard Medical School, Health Care Policy; Beth Israel Deaconess Medical Center, General Internal Medicine and Primary Care
Keywords: Symptom checkers, Internet, Antibiotic prescribing, mHealth
https://mc.manuscriptcentral.com/bmj
BMJ
Confidential: For Review O
nly
1
The Effectiveness of Symptom Checkers for Self-Diagnosis and Triage:
An Audit Study
Hannah L. Semigran,1
Jeffrey A. Linder,2 Courtney Gidengil,
3,4 Ateev Mehrotra,
1,5
1 Department of Health Care Policy, Harvard Medical School, Boston, MA
2 Division of General Medicine and Primary Care, Brigham and Women’s Hospital, Boston, MA
3 RAND Corporation, Boston, MA
4 Division of Infectious Diseases, Boston Children’s Hospital, Boston, MA
5 Division of General Internal Medicine and Primary Care, Beth Israel Deaconess Medical Center,
Boston, MA
Corresponding Author:
Ateev Mehrotra, MD, MPH
Harvard Medical School
180 Longwood Avenue
Boston, MA 02115
617-432-3905
Word Count: 3,689
Tables: 4
This study was funded by the United States’ National Institute of Health, (National Institute of Allergy
and Infectious Disease - Grant # R21 AI097759-01).
Contributors: Author Contributions: Study concept and design: Mehrotra, Semigran, Gidengil, Linder.
Acquisition of data: Semigran. Analysis and interpretation of data: Semigran, Mehrotra. Drafting of the
manuscript: Semigran. Critical revision of the manuscript for important intellectual content: Gidengil,
Linder, Mehrotra. Statistical analysis: Semigran, Mehrotra. Administrative, technical, and material
support: Mehrotra. Study supervision: Mehrotra
Page 1 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
2
Competing interests: All authors are affiliated with Harvard Medical School. Harvard Medical School’s
Family Health Guide is used as the basis for one of the symptom checkers evaluated. None of the
authors have been or plan to be involved in the development, evaluation, promotion or any other facet
of Harvard Medical School-related symptom checker. All authors have completed the ICMJE uniform
disclosure form at www.icmje.org/coi_disclosure.pdf and declare: no support from any organisation for
the submitted work; no financial relationships with any organisations that might have an interest in the
submitted work in the previous three years; no other relationships or activities that could appear to
have influenced the submitted work.
Ethical Approval: Not required
Data Sharing: No additional data available
Transparency: The senior author (the manuscript's guarantor) affirms that the manuscript is an honest,
accurate, and transparent account of the study being reported; that no important aspects of the study
have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered)
have been explained.
Page 2 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
3
SUMMARY:
Objectives: Use of the internet for self-diagnosis is common. Symptom checkers are online tools which
use computer algorithms to help patients with self-diagnosis and/or self-triage. Despite the growth in
use of such tools, the clinical accuracy of symptom checkers has not been assessed. Our objective was
to determine the diagnostic and triage accuracy of online symptom checkers.
Design: Audit study of web-based symptom checkers using standardized patient (SP) vignettes to assess
their clinical performance
Participants: We identified 23 symptom checkers that were in English, free, publicly-available, and
provided advice across a range of conditions. We compiled 45 standardized patient (SP) vignettes
equally divided into three categories of triage urgency: emergent care required (e.g., pulmonary
embolism), non-emergent care reasonable (e.g., otitis media), and self-care reasonable (e.g., viral upper
respiratory illness)
Main outcome measures: For symptom checkers that provided a diagnosis, our main outcomes were
whether the symptom checker listed the correct diagnosis first or within the first 20 potential diagnoses
(n = 770 SP evaluations). For symptom checkers that provided a triage recommendation, our main
outcomes were whether the symptom checker recommended emergent care, non-emergent care, or
self-care (n = 532 SP evaluations).
Results: The 23 symptom checkers provided the correct diagnosis first in 34% (95% confidence interval
[CI], 31 to 37) of SP evaluations, listed the correct diagnosis within the top 20 diagnoses given in 58%
(95% CI 55 to 62) of SP evaluations, and provided the appropriate triage advice in 57% (95% CI 52 to 61)
of SP evaluations. Triage performance varied by condition urgency, with appropriate triage advice
provided in 80% (95% CI 75 to 86) of emergent cases, 55% (95% CI 47 to 63) of non-emergent cases, and
Page 3 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
4
33% (95% CI 26 to 40) of self-care cases (p<0.001). Performance across the 23 individual symptom
checkers on appropriate triage advice ranged from 33% (95% CI 19 to 48) to 78% (95% CI 64 to 91) of SP
evaluations.
Conclusions: Symptom checkers had deficits in both triage and diagnosis. Triage advice from symptom
checkers is generally risk-averse, encouraging users to seek care for conditions where self-care is
reasonable.
Page 4 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
5
The public is increasingly using the internet to research their health concerns. For example, the United
Kingdom’s online patient portal for national health information, NHS Choices, reports over 15 million
visits per month.1 More than a third of adults in the United States regularly use the internet to self-
diagnose what ails them, using it for both non-urgent and urgent symptoms such as chest pain.2 3
While
there is a wealth of online resources to learn about specific conditions, self-diagnosis usually starts with
search engines like Google, Bing, or Yahoo.2 However, internet search engines can lead users to
confusing and sometimes unsubstantiated information and people with urgent symptoms are not
directed to seek emergent care.3-6
Recently, there has been a proliferation of more sophisticated
programs called symptom checkers that attempt to more effectively diagnose patients and direct them
to the appropriate care setting.3 6-13
Using computerized algorithms, symptom checkers ask users a series of questions about their symptoms
or require users to input their symptoms themselves. The algorithms vary and may use branching logic,
Bayesian inference, or other methods. Private companies and other organizations, including the
National Health Service (NHS), the American Academy of Pediatrics, and the Mayo Clinic, have launched
their own symptom checkers. One symptom checker, iTriage, reports 50 million uses per year.14
Typically symptom checkers are accessed via websites, but some are also available as apps for smart
phones or tablets.
Symptom checkers serve two main functions: facilitating self-diagnosis and assisting with triage. The
self-diagnosis function provides a list of diagnoses usually rank ordered by likelihood. The diagnosis
function is typically framed as helping educate patients on the range of diagnoses that might fit their
symptoms. The triage function informs patients whether they should seek care at all and, if so, where
(i.e. accident & emergency department, general practitioner’s (GP) clinic) and with what urgency (i.e.
emergently or within a few days). Symptom checkers may supplement or replace telephone triage lines,
Page 5 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
6
which are common in primary care.15-18
To ensure the safety of medical mobile apps, the United States
Congress is considering regulation of apps that “provide a list of possible medical conditions and advice
on when to consult a health care provider.”19 20
Symptom checkers have several potential benefits. They can encourage patients with a life-threatening
problem such as stroke or heart attack to seek emergency care.21
For patients with a non-emergent
problem that does not require a medical visit, these programs can reassure patients and recommend
they stay home. For approximately a quarter of acute respiratory illness visits for common conditions
like viral upper respiratory illness, patients do not receive any intervention beyond over-the-counter
therapy,22
and over half of patients receive unnecessary antibiotics.23-25
Decreasing unnecessary visits
saves patients’ time and money, deters overprescribing of antibiotics, and may decrease GP visits – a
critical issue given GP workload in the United Kingdom increased by 62% from 1995 to 2008.17
However,
there are several key concerns. If patients with a life-threatening problem are misdiagnosed and told to
not seek care, their health could worsen, increasing morbidity and mortality. Alternatively, if patients
with minor illnesses are told to seek care, in particular in an accident & emergency department, such
programs could increase unnecessary visits and therefore result in increased time and costs for patients
and society.
The impact of symptom checkers will depend to a large degree on their clinical performance, but to our
knowledge, no previous study has systematically evaluated the diagnostic and triage accuracy of
symptom checkers. To measure the accuracy of symptom checkers’ diagnosis and triage advice, we
audited 23 symptom checkers using 45 standardized patient (SP) vignettes. The vignettes reflected a
range of conditions from common to less common and low-acuity to life-threatening.
Methods
Search strategy for symptom checkers
Page 6 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
7
To identify symptom checkers, we searched for symptoms checkers between June 2014 and November
2014 that were in English, free, publicly-available, for humans (vs. veterinary), and did not focus on a
single type of condition (e.g. only orthopedic problems). To find symptom checkers that were available
as apps in the Apple App Store and Google Play, we used two search phrases (“symptom checker”,
“medical diagnosis”) used in a recent study on symptom checkers and examined by hand the first 240
search results.12
We used 240 because it has been used in prior studies that have searched app stores.26
To find online symptom checkers, we entered the same two search phrases in Google and Google
Scholar and examined the first 300 results. In prior research, the probability of relevant search results
declines substantially after the first 300 results.27
We supplemented our searches by asking two
symptom checker developers if they knew of other competing products.
In total, we identified 143 symptom checkers. We excluded 102 symptom checkers that used the same
medical content and logic as another tool (and therefore would have identical performance) (List
provided in Appendix). We excluded 25 symptom checkers that focused only on a single class of illness
(e.g. orthopedic problems), 14 that only provided medical advice (e.g. what symptoms are typically
associated with a certain condition) and did not provide diagnosis or triage advice, and 2 that were not
working. After these exclusions, we evaluated 23 symptom checkers.
Symptom checkers’ characteristics
We categorized symptom checkers by whether they facilitated self-diagnosis, self-triage, or both; type of
organization that operated the symptom checker; the maximum number of diagnoses provided, and
whether they were based on Schmitt or Thompson nurse triage guidelines, which are decision support
protocols commonly used in telephone triage for pediatric and adult consultations, respectively.28 29
We
grouped government and health plans together because both may have a financial incentive to deter
Page 7 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
8
unnecessary visits. We provide data when available about estimated total visitors to select symptom
checkers in the Appendix.
Clinical vignettes
To evaluate symptom checkers’ diagnosis and triage performance, we used 45 SP vignettes. We used
clinical vignettes to test performance because they are a common method to test physicians and other
clinicians on their diagnostic ability and management decisions. We purposefully selected SP vignettes
from 3 categories of triage urgency: (1) 15 SP vignettes for which emergent care is required, (2) 15 SP
vignettes for which non-emergent care is reasonable, and (3) 15 SP vignettes for which a medical visit is
generally unnecessary and self-care is sufficient. We chose vignettes across the severity spectrum
because patients are using symptom checkers for symptoms that require urgent and non-urgent care.3
We included vignettes for both common and uncommon conditions because we felt the clinical
community would be particularly interested in performance for less common but potentially life
threatening problems.
The SP vignettes were identified from various clinical sources, which also provided the associated
correct diagnosis. Symptom checkers ask users for a list of symptoms or ask a series of questions. Each
SP vignette was simplified into a core set of symptoms for easy entry and in some situations, we
supplemented the data provided by the vignette because a symptom checker asked about a symptom
not addressed in the vignette (details on source, core symptoms, and supplemental symptoms for each
vignette provided in Appendix).
We categorized the 45 vignette diagnosed as either “common” or “uncommon” diagnoses based on the
prevalence of the diagnosis among ambulatory visits in the United States (full details in Appendix).30
Assessing diagnosis and triage results
Page 8 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
9
Each SP vignette was entered into each website or app, and we recorded the resulting diagnoses and
triage advice. All vignettes were entered by an author (H.S.) who does not have clinical training. A
random sample of 25 SP vignettes were entered into symptom checkers by another person without
clinical training and there was high inter-rater reliability between the two in capturing the symptom
checker’s recommendations for diagnosis and triage (Cohen’s kappa 0.90). In some cases, we could not
evaluate a vignette because some symptom checkers only focus on children or adults or the symptom
checker did not list or ask for the key symptom in vignette. To avoid penalizing these symptom checkers,
we referred to SP vignettes that successfully yielded an output as “SP evaluations.”
To assess diagnostic accuracy, we noted whether the correct diagnosis was listed first or listed at all. For
several vignettes two symptom checkers presented a large number of diagnoses (up to 99). Because
such a long list of potential diagnoses is unlikely to be useful for a patient, we considered a diagnosis to
be listed at all only if it was within the first 20 diagnoses a symptom checker provided. It is possible that
many patients will only focus on the top diagnoses listed. Therefore, we also looked at whether the
correct diagnosis was listed in the first 3 diagnoses given. We judged the diagnosis incorrect if the
symptom checker indicated that the condition could not be identified.
We categorized the triage advice into three groups: (1) Emergent, which included advice to call an
ambulance, go to the accident & emergency department, or see your GP immediately; (2) Non-
Emergent, which included advice to call your GP or primary care provider, to see your GP or primary
care provider, go to an urgent care facility, go to a specialist, go to a retail clinic, or have an e-visit; and
(3) Self-Care, which included advice to stay at home or go to a pharmacy. If multiple triage locations
were suggested (e.g. accident & emergency department or a specialist), the most urgent suggestion was
used. We chose to do so because in almost all of the cases, the most urgent triage suggestion was listed
first. If a symptom checker was unable to reach a diagnosis decision for a given SP vignette but provided
Page 9 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
10
triage advice, we still assessed the appropriateness of this triage advice. Symptom checkers that
required users to select the correct diagnosis before giving triage advice were not included in assessing
triage accuracy with the exception of iTriage, which always suggested emergent triage advice.
Patient Involvement
There was no patient involvement in this study.
Analysis
We calculated summary statistics for diagnostic accuracy and triage advice with 95% confidence
intervals based on binomial distribution using Stata/MP 13.0. Given our focus on symptom checkers as a
whole, we did not make statistical comparisons of accuracy between individual symptom checkers. We
used chi-square tests to compare the diagnosis and triage accuracy by level and urgency and by type of
symptom checker. We conducted a sensitivity analysis of triage advice excluding several symptom
checkers that always or usually recommended emergent care.
Results
Study sample
The 23 identified symptom checkers were based in the United Kingdom, United States, Netherlands, and
Poland (Table 1): 11 symptom checkers provided both diagnoses and triage advice, 8 provided only
diagnoses, and 4 provided only triage advice. The 45 SP vignettes included 26 common and 19
uncommon diagnoses. Performance was assessed on a total of 770 SP evaluations for diagnosis and 532
SP evaluations for triage. Across the symptom checkers, 10 did not ask for demographics (age and
gender).
Accuracy of diagnosis
Page 10 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
11
Overall, the correct diagnosis was listed first in 34% (95% CI 31 to 37; Table 3) of SP evaluations.
Performance varied by urgency of condition. The correct diagnosis was listed first for 24% (95% CI 19 to
30) of emergent SP evaluations, 38% (95% CI 32 to 34) of non-emergent SP evaluations, and 40% (95% CI
34 to 47) of self-care SP evaluations (p<0.001 for comparison) (Table 3). There was no difference
between symptom checkers that did and did not ask for demographic information (34% [95% CI 30 to
39] vs. 34% [95% CI 28-39], p=0.88; Table 4). However, the correct diagnosis was listed first in SP
evaluations more often for common diagnoses compared to uncommon diagnoses (38% [95% CI 34 to
43] vs. 28% [95% CI 23 to 33], p = 0.004; Table 3).
Performance varied across symptom checkers (Table 2). Listing the correct diagnosis first in SP
evaluations ranged from 5% for MEDoctor (95% CI 0 to 13) to 50% for Doc Response (95% CI 33 to 67).
Few differences were observed by symptom checker characteristics (Table 5).
Across all symptom checkers the correct diagnosis was listed in the first 3 diagnoses in 51% (95% CI 47
to 54) of SP evaluations and in the first 20 diagnoses in 58% (95% CI 55 to 62) of SP evaluations (Table 3).
Diagnostic accuracy for listing the correct diagnosis in the top 3 and top 20 was higher for self-care
conditions vs. emergent conditions and was also higher for common conditions vs. uncommon
conditions. There was no significant difference in listing the correct diagnosis in the top 20 between
symptom checkers that listed more than 11 diagnoses versus those that only listed 1-3 diagnoses (59%
[95% CI 53 to 70] vs. 53% [95% CI 46 to 59], p=0.12; Table 4]. The accuracy of listing the correct diagnosis
in the top 20 across the 23 individual symptom checkers ranged from 34% (95% CI 17 to 52) to 84% (95%
CI 73 to 95) (Table 2).
Accuracy of triage advice
Appropriate triage advice was given in 57% (95% CI 52 to 61) of SP evaluations (Table 3). Performance
on triage advice was higher for emergent care SP evaluations versus non-emergent and self-care SP
Page 11 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
12
evaluations (80% [95% CI 75 to 86] vs. 55% [95% CI 47% to 63%] vs. 33% [95% CI 26 to 40], p<0.001).
Appropriate triage advice was higher for uncommon versus common diagnoses (63% [95% CI 57 to 70]
vs. 52% [95% CI 46 to 57], p=0.01).
iTriage, Isabel, and Healthwise advised emergent care for 100% of emergent SP evaluations. iTriage,
Symcat, Symptomate and Isabel always suggested users to seek care and therefore never advised self-
care (Table 2). After excluding these five symptom checkers, appropriate triage advice was 61% [95% CI
56 to 66]). (See Supplementary Table 5)
Symptom checkers that used the Schmitt or Thompson nurse-triage protocols were more likely to
provide appropriate triage decisions than those that did not (72% [95% CI 60 to 84] vs. 55% [95% CI 50
to 59], p=0.01; Table 4). Accurate triage advice varied by operator of symptom checker (provider groups
and physician associations 68% [95% CI 58 to 77], private companies 59% [95% CI 53 to 65], health
plans/governments 43% [95% CI 34 to 51], p<0.001).
Discussion
Principal Findings
Using SP vignettes we measured the diagnostic and triage accuracy of symptom checkers. Although
there was a range of performance across symptom checkers, overall they had deficits in both diagnosis
and triage accuracy. On average, symptom checkers provided the correct diagnosis within the first 20
listed in 58% of SP evaluations with the best-performing symptom checker listing the correct diagnosis
in 84% of SP evaluations. The correct triage decision was more than two times higher for SP evaluations
requiring emergent care (80%) versus those SP evaluations for which self-care was appropriate (34%).
Otherwise we found that symptom checkers advised the appropriate level of care about half the time,
but this varied by clinical severity.
Page 12 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
13
Comparisons with other studies
Our results on diagnostic accuracy and appropriate triage are roughly similar to prior work on the
performance of single symptom checkers for a limited set of diagnoses.6-8 31
An orthopedic symptom
checker listed the correct diagnosis for knee pain 89% of the time and Boots WebMD listed the correct
diagnosis 70% of the time for ear, nose, and throat symptoms.7 8
One study that also used two common
acute SP vignettes to evaluate WebMD reported a diagnostic accuracy rate of 50%.6
Whether this level of performance for diagnosis and triage is acceptable depends on the standard for
comparison. If symptom checkers are seen as a replacement for seeing a physician, they are likely an
inferior alternative. It is believed that physicians have a diagnostic accuracy rate of 85 to 90%, though in
some studies using clinical vignettes, performance was lower.32 33
However, in-person physician visits
might be the wrong comparison because patients are likely not using symptom checkers to obtain a
definitive diagnosis but for quick and accessible guidance. Also, instead of diagnostic accuracy the key
assessment of symptom checkers may be appropriate triage. Distinguishing between Rocky Mountain
spotted fever and meningitis may be less important than ensuring patients seek emergent care.
If symptom checkers are seen as an alternative for simply entering symptoms into an online search
engine such as Google, then symptom checkers are likely a superior alternative. A recent study found
that when typing acute symptoms that would require urgent medical attention into search engines to
identify symptom-related web sites, advice to seek emergent care was present only 64% of the time.3
Perhaps the most appropriate comparison to symptom checkers is telephone triage lines, which are
widely used in developed nations.15-18
In general patients use symptom checkers and telephone triage
for similar complaints.13
Also, many nurse phone triage lines use the same underlying clinical logic as the
symptom checkers evaluated in this study. For example, some health plan nurse triage lines use the
Healthwise symptom checker, and the Schmitt and Thompson protocols were originally developed for
Page 13 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
14
phone triage and now provide the underlying logic for several symptom checkers we evaluated. The
accuracy of telephone triage recommendations, as compared to in-person physician recommendations,
range from 61% in a study of pediatric abdominal pain to 69% in a multicenter observational study.34 35
A
recent study of NHS Symptom Checkers and NHS Direct’s telephone triage line found that triage advice
from both to be comparable.9 Given their similar clinical logic and performance in terms of triage,
symptom checkers may be viewed in general as a reasonable alternative to telephone triage. One
potential advantage of symptom checkers over telephone triage is cost.17
Because of their negligible
operation costs, symptom checkers could potentially be a more cost-effective way of providing triage
advice than nurse-staffed phone lines.
Implications for using symptom checkers
Both symptom checkers and telephone triage have been promoted as a means of reducing unnecessary
office visits.15-18 36
The impact of symptom checkers on how people seek care depends on how patients
respond to the advice provided which is unknown. In one study, users expressed skepticism about the
diagnosis ultimately suggested by a symptom checker.6 The risk-averse nature of symptom checker
triage advice is a concern. In two-thirds of SP evaluations where medical attention was not necessary,
we found symptom checkers encouraged care. Overly risk-adverse advice is not limited to symptom
checkers. Telephone triage advice can also encourage unnecessary care-seeking.31 34
For instance, the
NHS’s telephone triage line, which is not staffed by health professionals, has been implicated in
increasing accident & emergency room visits in the UK.37
Some patients researching health conditions
online are motivated by fear, and the listing of concerning diagnoses by symptom checkers could
contribute to hypochondriasis and “cyberchondria,” which describes the escalated anxiety associated
with self-diagnosis on the internet.38-42
Together, confusion, risk-adverse triage advice, and
cyberchondria could mean that symptom checkers encourage patients to receive care unnecessarily and
Page 14 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
15
thus increase health care spending. Understanding how patients interpret and use the symptom checker
advice and the impact of symptom checkers on care seeking should be a key focus for future research.
The symptom checkers in this study represent the first generation of such tools, and there is a number
of potential advances that may improve their performance in future versions. Incorporating local
epidemiological data may help inform diagnoses. For instance, addition of real-time information about
the local incidence of illness in the community greatly improved the performance of a Group A
Streptococcal pharyngitis diagnostic tool.10
Diagnosis and triage rates could also be improved if
symptom checkers incorporated population or individual clinical data from medical claims or the
electronic medical record. Demographic information is critical for both diagnostic and triage decisions
for programs like symptom checkers.11
One surprising finding in our study was that symptom checkers
that asked for demographic background information did not perform better. However, it is possible that
this demographic information was not effectively incorporated into the symptom checkers’ algorithms.
Strengths and limitations of this study
Despite the growth in use of symptom checkers, we believe ours is the first to assess clinical
performance across a large number of symptom checkers and wide range of conditions.
There were key limitations to this study. We cannot be sure we identified all publicly-available symptom
checkers, despite a thorough search of relevant databases and consultation with experts in this field. We
used clinical vignettes in which the symptoms and diagnoses were typically clear, and few had comorbid
conditions, resulting in a possible overestimation of the true clinical accuracy of symptom checkers.32
We also do not have data on the clinical performance of physicians for our SP vignettes, preventing a
direct comparison between symptom checkers and physicians beyond what is available in the literature.
When symptom checkers suggested several care sites (e.g. accident & emergency department or GP
Page 15 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
16
office), our triage assessment was based only on the highest acuity site of care listed and this may
contribute to our finding that triage advice is risk-averse.
Symptom checkers are part of a larger trend of both patients and physicians using the internet for many
health care tasks and therefore it appears likely that the use of symptom checkers will only increase.
Patients are chatting online with physicians,43
emailing their doctors for medical advice,44
receiving care
via E-Visits,45 46
and downloading health apps to smartphones.47
In addition to the public, physicians and
other practitioners are also using conceptually-similar tools to aid in the diagnosis and triage of their
patients.48 49
Physicians should be aware that an increasing number of their patients are using new internet-based
tools like symptom checkers, and that the diagnosis and triage advice patients receive may often be
inaccurate. For patients, our results imply that in many cases, symptom checkers can give the user a
sense of possible diagnoses but also provide a note of caution as the tools are frequently wrong and the
triage advice might be overly cautious. However, if the alternative to using a symptom checker is not
seeking any advice or simply using the internet, there may be value in their use. Further evaluations and
monitoring of symptom checkers will be important to assess whether they help people learn more and
make better decisions about their health.
Page 16 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
17
References
1. Gann B. Giving patients choice and control: health informatics on the patient journey. Yearb Med
Inform 2012;7(1):70-3.
2. Fox S, Duggan M. Health Online 2013. Internet and American Life Project. Washington D.C.: Pew
Research Center and California Health Care Foundation, 2013:4.
3. North F, Ward WJ, Varkey P, Tulledge-Scheitel SM. Should you search the Internet for information
about your acute symptom? Telemed J E Health 2012;18(3):213-8.
4. Black P. The dangers of using Google as a diagnostic aid. British Journal of Nursing 2009;18(19):1157.
5. Zhongbo C, Turner MR. The internet for self-diagnosis and prognostication in ALS. Amyotrophic
Lateral Sclerosis 2010;11:566.
6. Luger TM, Houston TK, Suls J. Older adult experience of online diagnosis: results from a scenario-
based think-aloud protocol. Journal of Medical Internet Research 2014;16(1):e16.
7. Bisson LJ, Komm JT, Bernas GA, Fineberg MS, Marzo JM, Rauh MA, et al. Accuracy of a computer-
based diagnostic program for ambulatory patients with knee pain. Am J Sports Med
2014;42(10):2371-6.
8. Farmer SE, Bernardotto M, Singh V. How good is Internet self-diagnosis of ENT symptoms using Boots
WebMD symptom checker? Clin Otolaryngol 2011;36(5):517-8.
9. Elliot AJ, Kara EO, Loveridge P, Bawa Z, Morbey RA, Moth M, et al. Internet-based remote health self-
checker symptom data as an adjuvant to a national syndromic surveillance system. Epidemiol
Infect 2015:1-7.
10. Fine AM, Nizet V, Mandl KD. Participatory medicine: A home score for streptococcal pharyngitis
enabled by real-time biosurveillance: a cohort study. Ann Intern Med 2013;159(9):577-83.
11. DocBot: A novel clinical decision support algorithm. Engineering in Medicine and Biology Society
(EMBC), 2014 36th Annual International Conference of the IEEE; 2014 26-30 Aug. 2014.
12. Lupton D, Jutel A. 'It's like having a physician in your pocket!' A critical analysis of self-diagnosis
smartphone apps. Soc Sci Med 2015;133:128-35.
13. North F, Varkey P, Laing B, Cha SS, Tulledge-Scheitel S. Are e-health web users looking for different
symptom information than callers to triage centers? Telemed J E Health 2011;17(1):19-24.
14. Reuters. Aetna Brings New iTriage Employer Technology to Mid-Sized Businesses, 2013.
15. Lattimer V, George S, Thompson F, Thomas E, Mullee M, Turnbull J, et al. Safety and effectiveness of
nurse telephone consultation in out of hours primary care: randomised controlled trial. The
South Wiltshire Out of Hours Project (SWOOP) Group. Bmj 1998;317(7165):1054-9.
16. Bunn F, Byrne G, Kendall S. The effects of telephone consultation and triage on healthcare use and
patient satisfaction: a systematic review. Br J Gen Pract 2005;55(521):956-61.
17. Campbell J, Fletcher E, Britten N, Green C, Holt T, Lattimer V, et al. Telephone triage for management
of same-day consultation requests in general practice (the ESTEEM trial): a cluster randomised
controlled trial and cost-sequence analysis. Lancet 2014;in press.
18. Richards D, Meakins J, Tawfik J, Godfrey L, Dutton E, Richardson G, et al. Nurse telephone triage for
same day appointments in general practice: multiple interrupted time series trial of effect on
workload and costs. British Medical Journal 2002;325(7374).
19. Lewis TL, Wyatt JC. mHealth and mobile medical Apps: a framework to assess risk and promote safer
use. Journal of Medical Internet Research 2014;16(9):e210.
20. Medical Electronic Data Technology Enhancement for Consumers' Health (MEDTECH). 2nd Session
ed, 2014.
21. Saczynski J, Yarzebski J, Lessard D, Spencer F, Gurwitz J, Gore J, et al. Trends in pre-hospital delay in
patients with acute myocardial infarction (from the Worcester Heart Attack Study). American
Journal of Cardiology 2008;102(12):1591.
Page 17 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
18
22. Barnett ML, Linder JA. Antibiotic prescribing for adults with acute bronchitis in the United States,
1996-2010. Jama 2014;311(19):2020-2.
23. Barnett ML, Linder JA. Antibiotic prescribing to adults with sore throat in the United States, 1997-
2010. JAMA Intern Med 2014;174(1):138-40.
24. Gonzales R, Malone DC, Maselli JH, Sande MA. Excessive antibiotic use for acute respiratory
infections in the United States. Clin Infect Dis 2001;33(6):757-62.
25. Little P, Rumsby K, Kelly J, Watson L, Moore M, Warner G, et al. Information leaflet and antibiotic
prescribing strategies for acute lower respiratory tract infection: a randomized controlled trial.
Jama 2005;293(24):3029-35.
26. Bierbrier R, Lo V, Wu RC. Evaluation of the accuracy of smartphone medical calculation apps. Journal
of Medical Internet Research 2014;16(2):e32.
27. Orizio G, Merla A, Schulz PJ, Gelatti U. Quality of online pharmacies and websites selling prescription
drugs: a systematic review. Journal of Medical Internet Research 2011;13(3):e74.
28. Schmitt BD. Pediatric Telephone Protocols: Office Version. Elk Grove Village IL: American Academy of
Pediatrics, 2012.
29. Thompson DA. Adult Telephone Protocols, 3rd Edition. Elk Grove Village IL: American Academy of
Pediatrics, 2013.
30. CDC, NAMCS, NHAMCS. Annual Number and Percent Distribution of Ambulatory Care Visits By
Setting Type According to Diagnosis Group: United States, 2009-2010, 2010.
31. Poote A, French D, Dale J, Powell J. A study of automated self-assessment in primary care student
health centre setting. Journal of Telemedicine and Telecare 2014;20(3):125.
32. Graber ML. The incidence of diagnostic error in medicine. BMJ Qual Saf 2013;22 Suppl 2:ii21-ii27.
33. Meyer AN, Payne VL, Meeks DW, Rao R, Singh H. Physicians' diagnostic accuracy, confidence, and
resource requests: a vignette study. JAMA Intern Med 2013;173(21):1952-8.
34. Staub GM, von Overbeck J, Blozik E. Teleconsultation in children with abdominal pain: a comparison
of physician triage recommendations and an established paediatric telephone triage protocol.
BMC Med Inform Decis Mak 2013;13:110.
35. Giesen P, Ferwerda R, Tijssen R, Mokkink H, Drijver R, van den Bosch W, et al. Safety of telephone
triage in general practitioner cooperatives: do triage nurses correctly estimate urgency? Qual
Saf Health Care 2007;16(3):181-4.
36. Stacey D, Graham I, O'Connor A, Pomey M. Barriers and facilitators influencing call center nurses'
decision support for callers facing values-sensitive decisions: A mixed methods study.
Worldviews on Evidence-Based Nursing 2005;2(4).
37. Donnelly L. A&E Crisis cause by NHS 111 phoneline, senior medic suggests. The Telegraph, January
14, 2015.
38. Usborne S. Cyberchondria: The perils of internet self-diagnosis. London: The Independent, February
17, 2009.
39. Hartzband P, Groopman J. Untangling the web - Patients, doctors, and the internet. New England
Journal of Medicine 2010;362(12):1064.
40. Brigo F, Igwe SC, Ausserer H, Nardone R, Tezzon F, Bongiovanni LG, et al. Why do people Google
epilepsy? An infodemiological study of online behavior for epilepsy-related search terms.
Epilepsy Behav 2014;31:67-70.
41. White RW, Horvitz E. Experiences with web search on medical concerns and self diagnosis. AMIA
Annu Symp Proc 2009;2009:696-700.
42. Husain I, Spence D. Can healthy people benefit from health apps? Bmj 2015;350:h1887.
43. Eminovic N, Wyatt J, Tarpey A, Murray G, Ingram G. First evaluation of the NHS Direct Online Clinical
Enquiry Service: A nurse-led web chat triage service for the public. Journal of Medical Internet
Research 2004;6(2).
Page 18 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
19
44. Eysenbach G, Diepgen T. Patients looking for information on the internet and seeking teleadvice.
Archives of Dermatology 1999;135(2).
45. Mehrotra A, Paone S, Martich GD, Albert SM, Shevchik GJ. A comparison of care at e-visits and
physician office visits for sinusitis and urinary tract infection. JAMA Intern Med 2013;173(1):72-
4.
46. DeJong C, Santa J, Dudley RA. Websites that offer care over the Internet: is there an access quality
tradeoff? Jama 2014;311(13):1287-8.
47. Edney A. The FDA Sets Its Sights on Medical Apps: Bloomberg Businessweek, September 20, 2013.
48. Cook DA, Enders F, Linderbaum JA, Zwart D, Lloyd FJ. Speed and accuracy of a point of care web-
based knowledge resource for clinicians: a controlled crossover trial. Interact J Med Res
2014;3(1):e7.
49. Sadeghi S, Barzi A, Sadeghi N, King B. A Bayesian model for triage decision support. Int J Med Inform
2006;75(5):403-11.
Page 19 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
20
Symptom
Checker Description
Maximum
No. of
Diagnoses*
Triage Options Provided
AskMD (USA) Online health and wellness platform from
Sharecare
(https://www.sharecare.com/askmd/get-
started)
15 n/a
BetterMedicine Health resource from HealthGrades;
symptom checker provides possible
diagnoses and information about these
conditions
(http://www.bettermedicine.com/symptom
-checker/)
46 n/a
DocResponse Symptom checker started by group of
certified physicians; user can choose from
internal medicine, dermatology, and
orthopedic views
(http://www.docresponse.com/)
5 n/a
Doctor
Diagnose
App offered on Google Play; provides
potential diagnoses and triage advice in
some cases
3 Seek immediate care, call your
doctor now, speak with your doctor,
home care
Drugs.com
(USA)
Online resource for drug and related health
information; uses content from Harvard
Health Publications
(http://www.drugs.com/symptom-
checker/)
10 ED, primary care doctor, home care
EarlyDoc
(Netherlands)
For triage criteria, uses Dutch College of
General Practitioners (NHG) TriageWijzer
and the Australian Triage Scale (used in
Australia and New Zealand to assess
urgency). (https://www.earlydoc.com/en/)
3 Don't wait and call a doctor now,
call a doctor preferably today, see
your doctor preferably on a
weekday, your complaints don't
seem urgent
Esagil (USA) Provides list of likely diagnoses (based on
the percent of entered symptoms that are
congruent with the diagnosis); the user can
also enter blood and urine analysis results
along with symptoms (http://esagil.org/)
65 n/a
Family Doctor
(USA)
Displays flow chart to track symptoms to a
diagnosis and triage option; produced by
the American Academy of Physicians
(http://familydoctor.org/familydoctor/en/h
ealth-tools/search-by-symptom.html)
7 ER, see your doctor, home care
FreeMD (USA) Takes user through a series of questions in a
"checkup" to finish with "what might be
wrong with you" and "where to go for
care"; owned by DSHI Systems, which
provides triage decision support solutions
from emergency medicine physicians to the
US government (Dep. of Veteran Affairs)
and private sector companies; program
called TriageXpert
(http://www.freemd.com/)
3 ED, urgent care, doctor's office,
doctor e-visit, retail clinic, dentist,
home care
Table 1|Symptom checkers included in the study
Page 20 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
21
Harvard
Medical School
Family Health
Guide (USA)
From Harvard Health Publications; this tool
is available both online and in print
(published 1999), and the online tool often
refers the user to the book in order to make
a diagnosis and triage decision
(http://www.health.harvard.edu/fhg/sympt
oms/symptoms.shtml)
4 ED, GP, home care
Healthline
(USA)
Health and wellness website that licenses
content to employers, health providers, and
health plans
(http://www.healthline.com/symptom-
checker)
76 n/a
Healthwise
(USA)
Non-profit provider for health content and
patient education; symptom checker
licensed to other organizations; we
accessed using the Province of Alberta's
website
(https://myhealth.alberta.ca/health/pages/
symptom-checker.aspx)
n/a Call 911 now, seek care now, seek
care today, try home care
Healthy
Children (USA)
From the American Academy of Pediatrics;
use's Barton D. Schmitt's "Pediatric
HouseCalls Symptom Checker" triage
protocol
(http://www.healthychildren.org/English/ti
ps-tools/symptom-checker)
n/a Call 911 now, call your doctor now
(night or day), call your doctor
within 24 hours, call your doctor
during weekday office hours, parent
care at home
Isabel (UK) Created by the Isabel Medical Charity
(http://symptomchecker.isabelhealthcare.c
om/suggest_diagnoses_advanced/landing_
page)
10 Walk in care, family doctor,
emergency services
iTriage (USA) Owned by Aetna; provides clinical sites in
user's region with addresses and phone
numbers (https://www.itriagehealth.com/)
5 Emergency department, urgent
care, retail clinic, family practice,
internal medicine, specialties,
prescription medication, over the
counter medication
Mayo Clinic
(USA)
Health resource website from Mayo Clinic
(http://www.mayoclinic.org/symptom-
checker/select-symptom/itt-20009075)
20 n/a
MEDoctor
(USA)
Free differential diagnosis system from
MEDoctor, Inc.
(https://www.medoctor.com/)
3 n/a
NHS Symptom
Checkers (UK)
Available through England's National Health
Services (NHS) Choices website
(https://www.nhs.uk/symptomcheckers/pa
ges/symptoms.aspx)
n/a Emergency department, general
practitioner, home care
Steps2Care
(USA)
iPhone and Android app, provides symptom
care guides from Barton D. Schmitt's
pediatric telephone triage guidelines and
David A. Thompson's adult telephone triage
guidelines
(http://www.paramounthealthcare.com/ste
ps2care )
n/a Call 911 now, go to ER now, Call
doctor now or go to ER, Call doctor
within 24 hours, Call doctor during
office hours, self-care at home
Symcat (USA) Triage tool uses data linking specific patient 6 Primary care, retail clinic,
Page 21 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
22
*Symptom checkers that provided diagnostic advice presented a list of potential diagnoses. We
identified the maximum number of diagnoses presented across the applicable vignettes.
symptoms and physician diagnoses across
visits seen in the NAMCS survey
(http://www.symcat.com/)
emergency room, urgent care
Symptify (USA) Online self-assessment tool and other
health services including emergency contact
list, consultation list etc.
(https://symptify.com/)
9 ER, urgent care, home care,
inconclusive
Symptomate
(Poland)
Uses Bayesian network methodology and a
medical database for diagnoses
(https://symptomate.com/)
5 ER, specialist, GP
WebMD (USA) Medical reference and health care resource
website (http://symptoms.webmd.com)
99 n/a
Page 22 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
23
Diagnosis
Triage
Symptom Checker
(n=23) Listed First Listed in Top 3 Listed in Top 20
All cases Emergent care required
Non-emergent care
reasonable Self-care reasonable
Rate* % (95%CI) Rate* % (95%CI) Rate* % (95%CI) Rate* % (95%CI) Rate* % (95%CI) Rate* % (95%CI) Rate* % (95%CI)
Ask MD 17/40 43 (26 to 59) 27/40 68 (52 to 83) 30/40 75 (61 to 89) - b
- b
- b -
b -
b -
b -
b -
b
BetterMedicine 11/45 24 (11 to 38) 13/45 29 (15 to 43) 17/45 38 (23 to 53) - b -
b -
b -
b -
b -
b -
b -
b
DocResponse 18/36 50 (33 to 67) 24/36 67 (50 to 83) 26/36 72 (57 to 88) - b -
b -
b -
b -
b -
b -
b -
b
Doctor Diagnose 16/39 41 (25 to 57) 17/39 44 (27 to 60) 18/39 46 (30 to 63) 10/16 63 (36 to 89) 8/10 80 (13 to 100) 2/3 67 (0 to 100) 0/3 0 (0 to 0)
Drugs.com 17/43 40 (24 to 55) 20/43 47 (31 to 63) 25/43 58 (43 to 74) 25/42 60 (44 to 75) 8/14 57 (27 to 87) 9/15 60 (32 to 88) 8/13 62 (31 to 92)
EarlyDoc 6/19 32 (9 to 55) 7/19 33 (9 to 57) 7/19 33 (9 to 57) 9/17 53 (26 to 79) 4/6 67 (12 to 100) 3/5 60 (0 to 100) 2/6 33 (0 to 88)
Esagil 9/44 20 (8 to 33) 15/44 34 (24 to 54) 22/44 50 (35 to 65) - b -
b -
b -
b -
b -
b -
b -
b
Family Doctor 20/43 47 (31 to 62) 24/43 56 (40 to 41) 24/43 56 (40 to 71) 22/41 54 (38 to 70) 6/12 50 (17 to 83) 7/14 50 (20 to 80) 9/15 60 (32 to 88)
FreeMD 16/44 36 (22 to 51) 20/44 45 (30 to 61) 21/44 48 (32 to 63) 26/44 59 (44 to 74) 10/15 67 (40 to 94) 13/15 87 (67 to 100) 3/14 21 (0 to 46)
HMS Family Health
Guide
13/38 34 (18 to 50) 20/38 52 (39 to 72) 21/38 55 (39 to 72) 32/40 78 (64 to 91) 13/14 93 (77 to 100) 11/13 85 (62 to 100) 8/13 62 (31 to 92)
Healthline 17/45 38 (23 to 53) 24/45 53 (37 to 68) 26/45 58 (43 to 73) - b -
b -
b -
b -
b -
b -
b -
b
Healthwise - a -
a -
a -
a -
a -
a 19/44 43 (28 to 58) 15/15 100 (100 to 100) 1/15 7 (0 to 21) 3/14 21 (0 to 46)
Healthy Children - a -
a -
a -
a -
a -
a 11/15 73 (48 to 99) 3/3 100 (100 to 100) 5/5 100 (100 to 100) 3/7 43 (0 to 92)
Isabel 20/45 44 (29 to 60) 31/45 69 (55 to 83) 38/45 84 (73 to 95) 23/45 51 (36 to 66) 15/15 100 (100 to 100) 8/15 53 (25 to 82) 0/15 0 (0 to 0)
iTriage 16/45 36 (22 to 51) 29/45 64 (49 to 78) 34/45 77 (64 to 90) 14/43 33 (19 to 48) 14/14 100 (100 to 100) 0/14 0 (0 to 0) 0/15 0 (0 to 0)
Mayo Clinic 7/41 17 (5 to 29) 24/41 59 (43 to 74) 31/41 76 (62 to 89) - b -
b -
b -
b -
b -
b -
b -
b
MEDoctor 2/37 5 (0 to 13) 16/37 43 (26 to 60) 16/37 43 (26 to 60) - b -
b -
b -
b -
b -
b -
b -
b
NHS - a -
a -
a -
a -
a -
a 23/44 52 (37 to 68) 13/15 87 (67 to 100) 3/15 20 (0 to 43) 7/14 50 (20 to 80)
Steps2Care - a -
a -
a -
a -
a -
a 30/42 71 (57 to 86) 12/14 86 (65 to 100) 10/14 71 (44 to 98) 8/14 57 (27 to 87)
Symcat 18/45 40 (25 to 55) 32/45 71 (57 to 85) 34/45 76 (62 to 89) 20/45 44 (29 to 60) 8/15 53 (25 to 82) 12/15 80 (57 to 100) 0/15 0 (0 to 0)
Symptify 13/45 29 (15 to 43) 16/45 36 (22 to 51) 20/45 44 (29 to 60) 28/40 70 (55 to 85) 11/12 92 (73 to 100) 10/14 71 (44 to 98) 7/14 50 (20 to 80)
Symptomate 10/32 31 (14 to 48) 11/32 34 (17 to 52) 11/32 34 (17 to 52) 9/14 64 (36 to 93) 7/9 76 (44 to 100) 2/3 67 (0 to 100) 0/2 0 (0 to 0)
WebMD 16/45 36 (21 to 50) 23/45 51 (36 to 66) 28/45 62 (47 to 77) - b -
b -
b -
b -
b -
b -
b -
b
Table 2|Accuracy of diagnosis decision and triage advice for each symptom checker
Abbreviations: HMS, Harvard Medical School; NHS, National Health Services
* Number of correct SP evaluations / Applicable SP evaluations. As noted in text, some SP vignettes could not be applied to a given symptom checker.
For example, we could not evaluate an adult SP vignette if it was a pediatric symptom checker. a Symptom checker does not provide diagnosis suggestions
b Symptom checker does not provide triage advice
Page 23 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
24
Diagnosis
Listed First Listed in Top 3
Listed in Top 20 Triage
No. of
vignettes
(%) Rate* %(95% CI) P value Rate* %(95% CI) P value Rate* %(95% CI) P value Rate* %(95% CI) P value
All Vignettes 45 (100) 262/770 34 (31 to 37) - 394/770 51 (47 to 54) - 449/770 58 (55 to 62) - 301/532 57 (52 to 61) -
Type of SP vignette
Emergent 15 (33) 64/263 24 (19 to 30)
<0.001
104/263 40 (34 to 46)
<0.001
132/263 50 (44 to 56)
0.003
147/183 80 (75 to 86)
< 0.001 Non-emergent 15 (33) 98/260 38 (32 to 44) 148/260 57 (51 to 63) 157/260 60 (54 to 66) 96/175 55 (47 to 63)
Self-care 15 (33) 100/247 40 (34 to 47) 142/247 57 (51 to 63) 160/247 65 (59 to 71) 58/174 33 (26 to 40)
Type of diagnosisa
Common 26 (58) 174/457 38 (34 to 43)
0.004
254/457 56 (52 to 61)
<0.001
283/457 62 (57 to 66)
0.01
162/313 52 (46 to 57)
0.01 Uncommon 19 (42) 88/313 28 (23 to 33)
140/313 45 (38 to 49)
166/313 53 (47 to 59)
139/219 63 (57 to 70)
* Number of correct SP evaluations / Applicable SP evaluations. As noted in text, some SP vignettes could not be applied to a given symptom checker.
For example, we could not evaluate an adult SP vignette if it was a pediatric symptom checker.
a Based on the annual number of ambulatory care visits in the United States, 2009-2010 (See Appendix for further description)
Table 3|Accuracy of diagnosis decision and triage advice for all symptom checkers, stratified by severity of the SP vignette and
by frequency of the condition’s diagnosis
Page 24 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
25
Diagnosis
Listed First Listed in Top 20 Triage
All Symptom
Checkers
No. of
symptom
checkers (%)
Rate* % (95% CI) P value
Rate* % (95% CI) P value
No. of
symptom
checkers (%) Rate* % (95% CI) P value
All Symptom
Checkers 23 (100)
19 (100)
262/770 34 (31 to 37) -
449/770 58 (55 to 62) -
15 (100) 301/532 57 (52 to 61) -
Uses nurse-triage books (Schmitt or Thompson)?
Yes 2 (9) 0 (0)
0 - a
- a
0 - a
- a
2 (13) 41/57 72 (60 to 84) 0.01
No 21 (91) 19 (100)
262/770 34 (31 to 37)
449/770 58 (55 to 62)
13 (87) 260/475 55 (50 to 59)
Asks demographic questions?
Yes 14 (61) 12 (63)
157/458 34 (30 to 39) 0.88
275/458 60 (56 to 65) 0.24
9 (60) 162/319 51 (45 to 56) <0.001
No 9 (39) 7 (37)
105/312 34 (28 to 39)
174/312 56 (50 to 61)
6 (40) 139/213 65 (59 to 72)
Site owner
Health plan or
government 3 (13)
1 (5)
16/44 36 (22 to 51)
0.9
34/44 77 (64 to 90)
0.01
3 (20) 56/131 43 (34 to 51)
< 0.001 Provider
group 4 (17)
5 (26)
56/167 34 (26 to 41)
104/167 62 (55 to 70)
4 (27) 65/96 68 (58 to 77)
Private
company 14 (61)
13 (68)
190/559 34 (30 to 38)
311/559 56 (52 to 60)
8 (53) 180/305 59 (53 to 65)
Maximum number of diagnoses listed
1-3 6 (32) 6 (32)
78/227 34 (28 to 41)
0.13
120/227 53 (46 to 59)
0.12
5 (33) 87/163 53 (46 to 60)
0.32 4-10 7 (37) 7 (37)
107/283 38 (32 to 43) 175/283 62 (56 to 68) 6 (40) 131/224 58 (52 to 65)
11+ 6 (32) 6 (32)
77/260 30 (24 to 35) 154/260 59 (53 to 65) 4 (27) - b -
b
* Number of correct SP evaluations / Applicable SP evaluations. As noted in text, some SP vignettes could not be applied to a given symptom
checker. For example, we could not evaluate an adult SP vignette if it was a pediatric symptom checker.
a Symptom checker does not provide diagnosis suggestions
b Symptom checker does not provide triage advice
Table 4|Accuracy of diagnosis given and triage advice for all symptom checkers given certain characteristics of the tools
Page 25 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nlyAppendix
This appendix includes more details on study sample exclusion criteria, additional methods, the
standardized patient (SP) vignettes that were used, diagnosis and triage accuracy for each symptom
checker, and the results of our sensitivity analyses.
Supplemental Table 1 categorizes the symptom checkers that were excluded from our study. After
identifying symptom checkers through the inclusion criteria described in the Methods, the symptom
checkers in this table were excluded on the basis of having the same underlying algorithm as another
tool in our sample or for other characteristics that we decided detracted from the ability of the
symptom checker to provide diagnostic and triage advice.
Our SP vignettes were gathered from several sources, which are listed in Supplemental Table 2. Each
vignette provided the age, gender, symptoms, and correct diagnosis for a given condition. This table also
notes where we added additional symptoms if the symptom checkers asked for them. Added symptoms
are italicized. The “simplified” symptoms were those inputted into each symptom checker.
To get a sense of the utilization of symptom checkers, we used Compete Pro to estimate the number of
unique visitors to symptom checker websites during the month of October 2014 in Supplemental Table
3. The limitations of this market analysis website, including its inability to track some websites outside of
the United States, those that were embedded within another website, and those with relatively low
traffic, allowed us to only estimate total use for seven symptom checkers.
Supplemental Table 4 has additional information for Table 2 in the manuscript. This includes the
accuracy of the diagnosis decision and triage advice for each symptom checker with the addition of the
stratification by the severity of the SP vignette.
Lastly, we performed sensitivity analyses shown in Supplemental Table 5 to assess the appropriateness
of the triage advice of the symptom checkers by excluding certain symptom checkers that were not as
variable in their triage advice. This includes iTriage, which always suggested that the user visit an
emergency department, and Symcat, Symptomate, and Isabel, all of which never suggest self-care.
Excluding these symptom checkers only had a modest impact on rates of appropriate triage advice.
Page 26 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nlyAdditional Methods
We stratified the performance of the symptom checkers by whether the diagnosis given by the SP
vignette was “common” or “uncommon.” We defined “common” diagnoses as those that accounted for
>0.3% of ambulatory visits (or >3,764,082 visits) in the United States in 2009-2010. These totals were
compiled from data gathered by the Center for Disease Control (CDC), the National Ambulatory Medical
Care Survey (NAMCS), and the National Hospital Ambulatory Medical Care Survey (NHAMCS).1
1. CDC, NAMCS, NHAMCS. Annual Number and Percent Distribution of
Ambulatory Care Visits By Setting Type According to Diagnosis Group:
United States, 2009-2010, 2010.
Page 27 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Supplemental Table 1: Symptom checkers excluded from study sample and reason for exclusion
Same Underlying Algorithm
Healthy Children (34) Steps2Care (28) iTriage (14)
Advocate Children’s Hospital AHN Health Finder Bayshore Community Hospital
Allied Pediatrics of New York Bon Secours Bryan Health
Children’s Medical Associates of Northern
Virginia Children’s Clinic of Raceland Crawford County Memorial Hospital
Children’s On Call College of Charleston HCA Far West
ChildrensMD Columbia St. Mary’s Inspira Health Network
ChildrensPGH East Tennessee Children’s Hospital Jersey Shore Medical Center
CIMG El Camino Hospital Meridian Health
COPA Eskenazi Health Mountainview Hospital
Docs2Go Indiana Univserity Health Ocean Medical Center
Greenwood Peditrics Intermountain health care OnPoint Urgent Care
HPN/SHL Lehigh Valley Health Network Riverside Community Hospital
Kid Aches Lourdes Hospital Riverview Medical Center
Kid Care St. Louis Children’s Mobile Middlesex Southern Ocean Medical Center
KidsDoc Mobile Nurse Sunrise Hospital and Medical Center
Lake Ray Hubbard Pediatrics Novant Health
MD 4Kids Pediatric Associate of Greater Salem NHS Symptom Checkers (6)
OU Medicine Providence Health and Services Health Direct Australia
PocketDoc Robert Wood Johnson University Hospital Martin Moth
REIS Pediatrics SCL Health NetDoctor
Sutter Health Mobile App Seton health care family NetDoctor
Swedish Kids South Texas Regional Medical Center North West Surrey
SymptomMD Spectrum Health Your.MD
UH Rainbow Babies and Children’s Hospital St. John Providence
Vanderbilt University Medical Center St. Vincent Health Healthwise (9)
Virtual Nurse UCLA Health Blue Shield of California
Page 28 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Wasatch Pediatrics Union County Hospital Group Health mobile
Wesley Kids UW Medicine Kaiser Permanente
Wesley Kids West Bloomfield Pediatrics Medical Mutual
Mercy
FreeMD (1)
Network of Care
EverydayHealth WebMD (2) Sutter Health website
MedicineNet The Hospital of Central Connecticut
Drugs.com (4) RxList University of Michigan Health System
Best Android Symptom Checker
GenieMD Isabel (2) Healthline (2)
King Abdullah bin Abdulaziz Arabic Health
Encyclopedia Patient.co.uk AARP
SmartHealth MSN Health and Fitness
Other Reasons for Exclusion
Tailored to specific condition (25) Symptom tracker (3) Medical advice only (14)
ADA Dental Symptom Checker Healee A.D.A.M. Symptom Checker
Capital Otolaryngology RheumaTrack About.com
Child Mind Institute Symple Alabama Blue Health Handbook
ColicCalm
Diagnosis And Therapy
Coping Cat Parents For pets (7) First Aid American Red Cross
Ebola Symptom Test Dog and Cat Dentist First Consult
First Aid and Symptom Checker PawNation How Stuff Works
Flu Alert Pet Education Medical Symptoms
Flu Facts PetCareRx Medical Wiz
Fortis Malar Hospital PetMD Parents.com
Hormone Balance Test WebDVM The Wellness Digest
MBH Symptom Checker ZooToo Urgent Care
MedZam Cold Flu
Xpress Urgent Care
MedZam Migraines Not working (2) Your Medial Encyclopedia
MedZam Restless Leg and Limb Dignity Health
Page 29 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
MedZam Strep Throat Saint Thomas Health
Meningitis
Myofascial Therapy.org Talk to a doctor (4)
Neocate Amwell
Pregnancy Test Doctor on Demand
Presbyterian/St. Luke’s MD Live
Shingles Symptom Checker RelyMD
SportsInjuryClinic
Trigger Point Products
USF Health
Page 30 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Diagnosis Vignette Simplified (added symptoms)
Acute liver failure¹ A 48-year-old woman with a history of migraine headaches presents to the emergency room with altered mental
status over the last several hours. She was found by her husband, earlier in the day, to be acutely disoriented and
increasingly somnolent. On physical examination, she has scleral icterus, mild right upper quadrant tenderness, and
asterixis. Preliminary laboratory studies are notable for a serum ALT of 6498 units/L, total bilirubin of 5.6 mg/dL, and
INR of 6.8. Her husband reports that she has consistently been taking pain medications and started taking additional
500 mg acetaminophen pills several days ago for lower back pain. Further history reveals a medication list with
multiple acetaminophen-containing preparations.
48 y/o f, confusion,
disorientation, increasingly
drowsy, mild right upper
quadrant tenderness, chronic
tylenol/acetaminophen -
recently took more
Appendicitis¹ A 12-year-old girl presents with sudden-onset severe generalized abdominal pain associated with nausea, vomiting,
and diarrhea. On exam she appears ill and has a temperature of 104°F (40°C). Her abdomen is tense with generalized
tenderness and guarding. No bowel sounds are present.
12 y/o f, sudden onset severe
abdominal pain, nausea,
vomiting, diarrhea, T=104
Asthma¹ A 27-year-old woman with a history of moderate persistent asthma presents to the emergency room with
progressive worsening of shortness of breath, wheezing, and cough over 3 days. She reports prior exposure to a
person who had a runny nose and a hacking cough. She did not receive significant relief from her rescue inhaler with
worsening symptoms, despite increased use. She has been compliant with her maintenance asthma regimen, which
consists of an inhaled corticosteroid and a leukotriene receptor antagonist for maintenance therapy and albuterol as
rescue therapy. Her cough is disrupting her sleep pattern and as a consequence she is experiencing daytime
somnolence, which is affecting her job performance.
27 y/o f, Hx of asthma, mild
shortness of breath,
wheezing, 3 days cough,
symptoms not responsive to
inhalers, recent cold
COPD flare (more severe)¹A 67-year-old woman with a history of COPD presents with 3 days of worsening dyspnea and increased frequency
of coughing. Her cough is now productive of green, purulent sputum. The patient has a 100-pack-year history of
smoking. She has had intermittent, low-grade fever of 100°F (37.7°C) for the past 3 days and her appetite is poor.
She has required increased use of rescue bronchodilator therapy in addition to her maintenance medications to
control symptoms.
67 y/o f, Hx of COPD, 3 days
worsening shortness of
breath, increase coughing,
green sputum, low grade
fever, increase use of rescue
bronchodilator therapy
Deep vein
thrombosis¹
A 65-year-old woman presents with unilateral leg pain and swelling of 5 days' duration. There is a history of
hypertension, mild CHF, and recent hospitalization for pneumonia. She had been recuperating at home but on
beginning to mobilize and walk, the right leg became painful, tender, and swollen. On examination, the right calf is 4
cm greater in circumference than the left when measured 10 cm below the tibial tuberosity. Superficial veins in the
leg are more dilated on the right foot and the right leg is slightly redder than the left. There is some tenderness on
palpation in the popliteal fossa behind the knee.
65 y/o f, 5 days swelling, pain
in one leg, recent
hospitalization, leg painful,
tender, swollen, red
Requires emergent care (n=15)
Supplemental Table 2: The 45 SP vignettes used to judge the symptom checkers’ accuracy and their condensed formats
Page 31 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Heart Attack² Mr. Y is a 64 year old Chinese male who presents with chest pain for 24 hours. One day prior to presentation, the
patient began to experience 8/10, non-radiating substernal chest pressure associated with diaphoresis and
shortness of breath. The pain intially improved with Tylenol, however over the following 24 hours, his symptoms
worsened. The patient went to his primary physician, where an EKG was performed which showed ST elevation in
leads V2-V6.
64 y/o m, 1 day chest pain
(8/10), non-radiating
substernal chest pressure,
sweating, shortness of
breath, (chest tightness )
Hemolytic uremic
syndrome¹
A 4-year-old boy presents with a 7-day history of abdominal pain and watery diarrhea that became bloody after the
first day. Three days before the onset of symptoms, he had visited the county fair with his family and had eaten a
hamburger. Physical examination reveals a mild anemia
4 y/o m, 7 day Hx of
abdominal pain, bloody
diarrhea, ate hamburger at
fair 3 days ago
Kidney stones¹ A 45-year-old white man presents to the emergency department with a 1-hour history of sudden onset of left-sided
flank pain radiating down toward his groin. The patient is writhing in pain, which is unrelieved by position. He also
complains of nausea and vomiting.
45 y/o m, 1 hour severe left-
sided flank pain radiating into
groin, nausea, vomiting, pain
unrelieved by position
Malaria¹ A 28-year-old man presents to his physician with a 5-day history of fever, chills, and rigors, not improving with
acetaminophen (paracetamol), along with diarrhea. He had been traveling in Central America for 3 months,
returning 8 weeks ago. He had been bitten by mosquitoes on multiple occasions, and although he initially took
malaria prophylaxis, he discontinued it due to mild nausea. He does not know the specifics of his prophylactic
therapy. On examination he has a temperature of 100.4°F (38°C), and is mildly tachycardic with a BP of 126/82
mmHg. The remainder of the examination is normal.
28 y/o m, 5 day Hx of fever,
chills, rigors, diarrhea, recent
travel abroad to area with
malaria, bitten by
mosquitoes, did not take
malaria prophylaxis
consistently
Meningitis¹ An 18-year-old male student presents with severe headache and fever that he has had for 3 days. Examination
reveals fever, photophobia, and neck stiffness.
18 y/o m, 3 days severe
headache, fever,
photophobia, neck stiffness
Pneumonia³ A 65-year-old man with hypertension and degenerative joint disease presents to the emergency department with a
three-day history of a productive cough and fever. He has a temperature of 38.3°C (101°F), a blood pressure of
144/92 mm Hg, a respiratory rate of 22 breaths per minute, a heart rate of 90 beats per minute, and oxygen
saturation of 92 percent while breathing room air. Physical examination reveals only crackles and egophony in the
right lower lung field. The white-cell count is 14,000 per cubic millimeter, and the results of routine chemical tests
are normal. A chest radiograph shows an infiltrate in the right lower lobe.
65 y/o m, Hx of hypertension
and degenerative joint
disease, 3 day Hx of
productive cough and fever
(101)
Page 32 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Pulmonary embolism¹ A 65-year-old man presents to the emergency department with acute onset of SOB of 30 minutes' duration. Initially,
he felt faint but did not lose consciousness. He is complaining of left-sided chest pain that worsens on deep
inspiration. He has no history of cardiopulmonary disease. A week ago he underwent a total left hip replacement
and, following discharge, was on bed rest for 3 days due to poorly controlled pain. He subsequently noticed swelling
in his left calf, which is tender on examination. His current vital signs reveal a fever of 100.4°F (38.0°C), heart rate
112 bpm, BP 95/65, and an O2 saturation on room air of 91%.
65 y/o m, shortness of breath
for 30 min, chest pain that
worsens with inspiration,
recent surgery, recent bed
rest, swelling in left calf,
which is tender, fever
Rocky Mountain
Spotted Fever4
An 8-year-old boy in Oklahoma is brought to the emergency department over the fourth of July weekend because of
fever, chills, malaise, athralgias, and a headache. Physical examination reveals a maculopapular rash that is most
prominent on his wrists and ankles.
8 y/o m, Fever, chills, joint
pain, headache, rash
wrists/ankles
Stroke¹ A 70-year-old man with a history of chronic HTN and atrial fibrillation is witnessed by a family member to have
nausea, vomiting, and right-sided weakness, as well as difficulty speaking and comprehending language. The
symptoms started with only mild slurred speech before progressing over several minutes to severe aphasia and
right arm paralysis. The patient is taking warfarin.
70 y/o m, nausea, vomiting,
right-sided weakness, rt arm
paralysis, difficulty speaking
and comprehension
Tetanus¹ A 63-year-old man sustained a cut on his hand while gardening. His immunization history is significant for not having
received a complete tetanus immunization schedule. He presents with signs of generalized tetanus with trismus
("lock jaw"), which results in a grimace described as "risus sardonicus" (sardonic smile). Intermittent tonic
contraction of his skeletal muscles causes intensely painful spasms, which last for minutes, during which he retains
consciousness. The spasms are triggered by external (noise, light, drafts, physical contact) or internal stimuli, and as
a result he is at the risk of sustaining fractures or developing rhabdomyolysis. The tetanic spasms also produce
opisthotonus, board-like abdominal wall rigidity, dysphagia, and apneic periods due to contraction of the thoracic
muscles and/or glottal or pharyngeal muscles. During a generalized spasm the patient arches his back, extends his
legs, flexes his arms in abduction, and clenches his fists. Apnea results during some of the spasms. Autonomic
overactivity initially manifests as irritability, restlessness, sweating, and tachycardia. Several days later this may
present as hyperpyrexia, cardiac arrhythmias, labile hypertension, or hypotension.
65 y/o m, cannot open
mouth, contraction of
muscles causing painful
spasms for minutes,
sweating, tachycardia, cut
hand while gardening, did not
get tetanus shot
Acute otitis media¹ An 18-month-old toddler presents with 1 week of rhinorrhea, cough, and congestion. Her parents report she is
irritable, sleeping restlessly, and not eating well. Overnight she developed a fever. She attends day care and both
parents smoke. On examination signs are found consistent with a viral respiratory infection including rhinorrhea and
congestion. The toddler appears irritable and apprehensive and has a fever. Otoscopy reveals a bulging,
erythematous tympanic membrane and absent landmarks.
18 mo f, 1 week rhinorrhea,
cough, congestion, irritable,
lack of appetite, fever, in
daycare
Requires non-emergent care (n=15)
Page 33 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Acute pharyngitis¹ A 7-year-old girl presents with abrupt onset of fever, nausea, vomiting, and sore throat. The child denies cough,
rhinorrhea, or nasal congestion. On physical exam, oral temperature is 101°F (38.5°C) and there is an exudative
pharyngitis, with enlarged cervical lymph nodes. A rapid antigen test is positive for group A Streptococcus (GAS).
7 y/o f, fever (101), nausea,
vomiting, sore throat, swollen
lymph nodes, tonsilar
exudate; no cough,
rhinorrhea, or nasal
congestion
Acute pharyngitis5 Mr. A is a 24 year-old man who presents to your office for complaints of sore throat, fever, and headache. His
symptoms started 2 days ago with acute onset of sore throat and fever to 102.2. He has had no cough. His physical
examination is normal, except for the presence of tonsillar exudates and some tender anterior cervical
lymphadenopathy. He is otherwise in good health, and is on no medications except for ibuprofen for fever. He has
no drug allergies. (, Centor score = 4 – treat, or test and treat)
24 y/o m, sore throat, fever
(102.2), headache, no
cough,tonsilar exudates
Acute sinusitis5 Mrs. S is a 35 year-old woman who presents with 15 days of nasal congestion. She has had facial pain and green
nasal discharge for the last 12 days. She has had no fever. On physical examination, she has no fever and the only
abnormal finding is maxillary tenderness on palpation. She is otherwise healthy, except for mild obesity. She is on no
medications, except for an over-the-counter decongestant. She has no drug allergies
35 y/o f, sx for 15 days, nasal
congestion, facial pain, green
nasal discharge, no fever
Back pain6 Consider a 35-year-old man who developed low back pain after shoveling snow 3 weeks ago. He presents to the
office for an evaluation. On examination there is a new left foot drop. In study 82% physicians recommend MRI
(sciatica/sprain)
35 y/o m, back pain following
shoveling, left foot drop,
symptoms 3 weeks of
duration (loss of sensation in
foot)
Cellulitis¹ A 45-year-old man presents with acute onset of pain and redness of the skin of his lower extremity. Low-grade fever
is present and the pretibial area is erythematous, edematous, and tender.
45 y/o m, pain and redness of
skin, low grade fever,
redness, edema, and
tenderness lower leg
COPD flare (milder)¹ A 56-year-old woman with a history of smoking presents to her primary care physician with shortness of breath and
cough for several days. Her symptoms began 3 days ago with rhinorrhea. She reports a chronic morning cough
productive of white sputum, which has increased over the past 2 days. She has had similar episodes each winter for
the past 4 years. She has smoked 1 to 2 packs of cigarettes per day for 40 years and continues to smoke. She denies
hemoptysis, chills, or weight loss and has not received any relief from over-the-counter cough preparations.
56 y/o f, Hx of smoking,
shortness of breath and
cough for several days,
rhinorrhea 3 days ago, white
sputum, no chills
Influenza¹ A 30-year-old woman presents in January with 2-day history of fever, cough, headache, and generalized weakness.
She was in her usual state of health before an abrupt onset of these symptoms. A few viral illnesses have affected
her during the current winter, but not to this severity. She reports sick contacts at work and did not receive the
seasonal influenza vaccine this season.
30 y/o f, 2 day fever, cough,
headache, weakness, did not
get flu shot
Page 34 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Mononucleosis¹ A 16-year-old female high school student presents with complaints of fever, sore throat, and fatigue. She started
feeling sick 1 week ago. Her symptoms are gradually getting worse, and she has difficulty swallowing. She has had a
fever every day, and she could hardly get out of bed this morning. She does not remember being exposed to
anybody with a similar illness recently. On physical examination she is febrile and looks sick. Enlarged cervical lymph
nodes, exudative pharyngitis with soft palate petechiae and faint erythematous macular rash on the trunk and arms
are found.
16 y/o f, 1 week Hx of fever,
sore throat, fatigue, difficulty
swallowing, fever, enlarged
lymph nodes, exudates,
macular rash on trunk/arms
Peptic Ulcer Disease¹ A 40-year-old man presents to his primary care physician with a 2-month history of intermittent upper abdominal
pain. He describes the pain as a dull, gnawing ache. The pain sometimes wakes him at night, is relieved by food and
drinking milk, and is helped partially by ranitidine. He had a similar but milder episode about 5 years ago, which was
treated with omeprazole. Physical examination reveals a fit, apparently healthy man in no distress. The only
abnormal finding is mild epigastric tenderness on palpation of the abdomen.
40 y/o m, 2 month Hx of
intermittent upper abdominal
pain, dulling and gnawing
ache, wakes at night and is
relieved by food/drinking
milk/ranitidine, prior episode
5 yrs ago
Pneumonia¹ A 6-year-old boy with a medical history significant for mild persistent asthma is brought to the clinic by his mother
with a history of a 5-day cough. His mother reports that the child's fever continues to be elevated despite
acetaminophen therapy. He has missed school for the past 3 days and he has a classmate sick with pneumonia. The
mother reports that the appetite is good for the child. His cough produced yellowish sputum at home. His vitals at
the clinic are: respiratory rate 19 breaths/min, heart rate 80 beats/min, and temperature 101.6°F (38.7°C). He
appears in no respiratory distress. His lung examination reveals bilateral rales and occasional wheeze. CXR reveals
lobar infiltrates without pleural effusions.
6 y/o m, Hx of asthma, 5 days
cough, fever, appetite good,
yellow sputum, t 101.6
Salmonella¹ A 14-year-old boy presents with nausea, vomiting, and diarrhea. Eighteen hours earlier, he had been at a picnic
where he ingested undercooked chicken along with a variety of other foods. He reports moderate-volume,
nonbloody stools occurring 6 times a day. He has mild abdominal cramps and a low-grade fever. He is evaluated at
an acute care clinic and found to be mildly tachycardic (heart rate 105 bpm) with a normal BP and a low-grade
temperature of 100.1°F (37.8°C). His physical exam is unremarkable except for mild diffuse abdominal tenderness
and mild increased bowel sounds. He is able to take oral fluids and is instructed on the appropriate oral fluid and
electrolyte rehydration.
14 y/o m, nausea, vomiting,
non-bloody diarrhea, mild
abdominal cramps (T=100.1),
mild abdominal tenderness,
diarrhea after attending a
picnic and eating
undercooked chicken,
Shingles¹ A 77-year-old man reports a 5-day history of burning and aching pain on the right side of his chest. This is followed
by the development of erythema and a maculopapular rash in this painful area, accompanied by headache and
malaise. The rash progressed to develop clusters of clear vesicles for 3 to 5 days, evolving through stages of
pustulation, ulceration, and crusting.
77 y/o m, 5 day burning and
aching on right side of chest,
erythema, maculopapular
rash, headache, malaise, rash
progressed to clear vesicles
after 3-5 days
Page 35 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Urinary tract
infection¹
A 26-year-old female newly wed presents complaining of painful urination, feeling of urgent need to urinate, and
more frequent urination for 2 days. She denies any fever, chills, nausea, vomiting, back pain, vaginal discharge, or
vaginal pruritus.
26 y/o f, painful urination,
urgent need to urinate, more
frequent urination for 2 days,
sexually active; no fever,
chills, nausea, vomiting, back
pain, vaginal discharge,
vaginal pruritus
Vertigo¹ A 65-year-old woman presents with a chief complaint of dizziness. She describes it as a sudden and severe spinning
sensation precipitated by rolling over in bed onto her right side. Symptoms typically last <30 seconds. They have
occurred nightly over the last month and occasionally during the day when she tilts her head back to look upward.
She describes no precipitating event prior to onset and no associated hearing loss, tinnitus, or other neurologic
symptoms. Otologic and neurologic examinations are normal except for the Dix-Hallpike maneuver, which is negative
on the left but strongly positive on the right side.
65 y/o f, dizziness, sudden
onset, recurrent, lasts <30
sec, consistent trigger, no
hearing loss, ringing in ears,
muscle weakness, loss of
sensation
Acute bronchitis¹ A 34-year-old woman with no known underlying lung disease 12-day history of cough. She initially had nasal
congestion and a mild sore throat, but now her symptoms are all related to a productive cough without paroxysms.
She denies any sick contacts. On physical examination she is not in respiratory distress and is afebrile with normal
vital signs. No signs of URI are noted. Scattered wheezes are present diffusely on lung auscultation.
34 y/o f, 12 day cough, initial
nasal congestion and sore
throat, cough, no fever
Acute bronchitis5 Mrs. L is a 61 year-old woman who presents with 4 days of a cough productive of yellow sputum. Her symptoms
started 4 days ago with rhinorrhea and productive cough. She initially had fevers as high as 101 for 2 days, but those
have now resolved. In the office, she has normal vital signs and a normal physical examination. She is otherwise
healthy except for high cholesterol for which she is being treated with atorvastatin. She has no drug allergies.
61 y/o f, 4 day cough, yellow
sputum, rhinorrhea, fever
(resolved)
Acute conjunctivitis¹ A 14-year-old boy with no significant past medical history presents 3 days after developing a red, irritated right eye
that spread to the left eye today. He has watery discharge from both eyes and they are stuck shut in the morning. He
reports recent upper respiratory symptoms and that several children at his day camp recently had pink eye. He
denies significant pain or light sensitivity and does not wear contact lenses. On examination, his pupils are equal and
reactive and he has a right-sided, tender preauricular lymph node. Penlight examination does not reveal any corneal
opacity.
14 y/o m, 3 days red, irritated
eye (spread from right to
left), discharge, URI
symptoms, no pain or light
sensitivity
Acute pharyngitis5 Mr. E is a 26 year-old man who presents to your office for complaints of sore throat, headache, and non-productive
cough. His symptoms started 2 days ago with acute onset of sore throat. He has been afebrile. His physical
examination is normal, except for some pharyngeal erythema. He is otherwise in good health, and is on no
medications except for acetaminophen for his sore throat and fever. He has no drug allergies.
26 y/o m, 2 day sore throat,
headache, cough, no fever
Self-care appropriate (n=15)
Page 36 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Allergic rhinitis¹ A 22-year-old student presents with a 5-year history of worsening nasal congestion, sneezing, and nasal itching.
Symptoms are year-round but worse during the spring season. On further questioning it is revealed that he has
significant eye itching, redness, and tearing as well as palate and throat itching during the spring season. He
remembers that his mother told him at some point that he used to have eczema in infancy.
22 y/o m, 5 year Hx of nasal
congestion, sneezing, nasal
itching worse during spring
season, eye itching, redness,
tearing, palate and throat
itching, Hx of eczema in
infancy
Back pain¹ A 38-year-old man with no significant history of back pain developed acute LBP when lifting boxes 2 weeks ago. The
pain is aching in nature, located in the left lumbar area, and associated with spasms. He describes previous similar
episodes several years ago, which resolved without seeing a doctor. He denies any leg pain or weakness. He also
denies fevers, chills, weight loss, and recent infections. Over-the-counter ibuprofen has helped somewhat, but he
has taken it only twice a day for the past 3 days because he does not want to become dependent on painkillers. On
examination, there is decreased lumbar flexion and extension secondary to pain, but a neurologic exam is
unremarkable.
38 y/o m, acute low back pain
after lifting, no leg pain or
weakness, no fevers, chills,
weight loss, or recent
infections
Bee sting without
anaphylaxis¹
A 9-year-old boy is brought to the ER after being stung by a bee at a picnic. He is crying hysterically. After 15 minutes
of calming him down, exam reveals a swollen tender upper lip but no tongue swelling, no drooling, no stridor, no
rash, and no other complaints.
9 y/o m, bee sting, swollen
and tender upper lip; no
tongue swelling, drooling,
stridor, rash, or other
complaints
Canker sore¹ A 17-year-old male student presents with recurrent mouth ulceration since his early schooldays. He has no
respiratory, anogenital, gastrointestinal, eye, or skin lesions. His mother had a similar history as a teenager. The
social history includes no tobacco use and virtually no alcohol consumption. He has no history of recent drug or
medication ingestion. Extraoral exam reveals no significant abnormalities and specifically no pyrexia; no cervical
lymph node enlargement; nor cranial nerve, salivary, or temporomandibular joint abnormalities. Oral exam reveals a
well-restored dentition and there is no clinical evidence of periodontal-attachment loss or pocketing. He has five 4
mm round ulcers with inflammatory haloes in his buccal mucosae.
17 y/o m with recurrent
mouth ulceration for year, no
respiratory, anogenital,
gastrointestinal, eye, or skin
lesions, mother has similar
Hx, no Hx of recent drugs or
medication
Candidal yeast
infection6
Consider a 40-year-old, monogamous, married woman who calls to report a 2-day history of vaginal itching and
thick white discharge. She has no abdominal pain or fever. (in study 50% recommended physician visit)
40 y/o f, 2 day vaginal itching,
thick white discharge, no
abdominal pain or fever
Page 37 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Constipation¹ A 5-month-old baby boy presents with difficulty and delay in passing hard stools. His mother reports that he strains
for several hours and may even miss a day, before passing stool with screaming and occasional spots of fresh blood
on the stool or diaper. He has recently been weaned from breastfeeding to cows' milk formula, which he had been
reluctant to drink initially. The child is thriving and now feeding normally. There was no neonatal delay in defecation
and no history of excessive vomiting or abdominal distension.
5 mo m, difficulty/delay in
passing hard stools, strains
for hours, may miss a day,
screams when passes stool
and occasional spots of
blood, weaned from
breastmilk to cows' milk, now
feeding normally
Eczema¹ A 12-year-old female presents with dry, itchy skin that involves the flexures in front of her elbows, behind her knees,
and in front of her ankles. Her cheeks also have patches of dry, scaly skin. She has symptoms of hay fever and has
recently been diagnosed with egg and milk allergy. She has a brother with asthma and an uncle and several cousins
who have been diagnosed with eczema.
12 y/o f, dry, itchy skin in
front of elbows, behind
knees, in front of ankles,
cheeks have patches of dry,
scaly skin, symptoms of hay
fever, egg and milk allergy,
brother has asthma and uncle
and cousins have eczema
Stye¹ A 30-year-old man presents with a painful, swollen right eye for the past day. He reports minor pain on palpation of
the eyelid and denies any history of trauma, crusting, or change in vision. He has no history of allergies or any eye
conditions and denies the use of any new soaps, lotions, or creams. On exam, he has localized tenderness to
palpation and erythema on the midline of the lower eyelid near the lid margin. The remainder of the physical exam,
including the globe, is normal.
30 y/o m, painful, swollen
right eye for past day, no Hx
of trauma, crusting, change in
vision, allergies, or eye
conditions, localized
tenderness, erythema
(redness)
Viral upper respiratory illnessMr. R. is a 56 year-old man who presents to you with 6 days of non-productive cough, nasal congestion, and green
nasal discharge. He has had intermittent fevers as high as 100.8. His physical examination is normal except for
rhinorrhea. He is otherwise healthy, except for chronic osteoarthritis of the right knee. He has no drug allergies.
56 y/o m, 6 day cough, nasal
congestion, green nasal
discharge, fever (100.8),
rhinorrhea
Page 38 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Viral upper respiratory illness¹A 30-year-old man presents with a 2-day history of runny nose and sore throat. He feels hot and sweaty, has a mild
headache, is coughing up clear sputum and complains of muscle aches. He would like antibiotics as he was
prescribed them last year when he had a similar condition. On examination, he is afebrile, has a normal pulse, a
slightly inflamed pharynx and nontender cervical lymphadenopathy. There is no neck stiffness and his chest is clear.
He has tried over-the-counter cough medications, but has not found these helpful. He smokes 10 cigarettes per day.
30 y/o m, 2 day HX of runny
nose, sore throat, hot,
sweaty, mild headache, cough
with clear sputum, muscle
aches, no fever or neck
stiffness
Vomiting7 Elizabeth’s 2-year-old son has a fever and vomited twice. Elizabeth worries about dehydration, so she gives Jack a
sippy cup of apple juice. He immediately vomits up the juice. Elizabeth debates what to do next. Should she try to
reach Jack’s pediatrician or should she take Jack to the ED? Instead, she calls her triage nurse line. Temperature =
100.5
2 y/o m, low grade fever (T =
100.5), vomited twice, vomits
up juice
Table References
1. Epocrates. https://online.epocrates.com/noFrame/.
2. Lue J. NYU Medical Grand Rounds Clinical Vignette. 2012;
http://www.medicine.med.nyu.edu/education/im-residency-homepage/research-
opportunities/clinical-vignettes. Accessed September 8, 2014.
3. Halm EA, Teirstein AS. Clinical practice. Management of community-acquired pneumonia. N Engl J
Med. Dec 19 2002;347(25):2039-2045.
4. Plantz SH, Adler JN, eds. NMS Emergency Medicine. Baltimore: Williams & Wilkins; 1998. National
Medical Series for Independent Study.
5. Gidengil CA, Linder J, Beach S, Setodjian C, Hunter G, Mehrotra A. Using clinical vignettes to predict
antibiotic prescribing for acute respiratory infections. In review.
6. Sirovich BE, Gottlieb DJ, Welch HG, Fisher ES. Variation in the tendency of primary care physicians
to intervene. Arch Intern Med. Oct 24 2005;165(19):2252-2256.
7. Boroughs DS, Dougherty JA, Goldsmith C. Telephone Triage: Help Is Just a Call Away.
http://ce.nurse.com/RVignette.aspx?TopicId=718. Accessed September 10, 2014.
Page 39 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Supplemental Table 3: Estimation of unique visitors to symptom checker websites in October 2014 using Compete Pro.¹
Symptom Checker Number of Unique Visitors
AskMD/Sharecare 4,220
EarlyDoc 16,826
FreeMD 37,545
Isabel 10,517
iTriage 198,398
Symcat 2,889
WebMD 975,127
Total 1,246,071
¹Compete. Site Comparison, 2014.
Page 40 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Do
cRe
spo
nse
Fa
mil
y D
oct
or
Isa
be
l
Ask
MD
/
Sh
are
care
Do
cto
r D
iag
no
se
Dru
gs.
com
Sy
mca
t
He
alt
hli
ne
Fre
eM
D
iTri
ag
e
We
bM
D
HM
S F
am
ily
He
alt
h G
uid
e
Ea
rly
Do
c
Sy
mp
tom
ate
Sy
mp
tify
Be
tte
r M
ed
icin
e
Esa
gil
Ma
yo
Cli
nic
ME
Do
cto
r
HM
S F
am
ily
He
alt
h G
uid
e
He
alt
hy
Ch
ild
ren
Ste
ps2
Ca
re
Sy
mp
tify
Sy
mp
tom
ate
Do
cto
r D
iag
no
se
Dru
gs.
com
Fre
eM
D
Fa
mil
y D
oct
or
Ea
rly
Do
c
NH
S
Isa
be
l
Sy
mca
t
He
alt
hw
ise
iTri
ag
e
Standardized Patient Vignette Emergent Care Non Emergent Care Self Care
Require emergent care
Acute liver failure 〇 〇 〇 〇 U
Appendicitis 〇 ⊗ ⊗ 〇 〇 〇 〇 ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ▨Asthma attack ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ 〇 〇 〇 ⊗ ⊗ ⊗
COPD more severe ⊗ 〇 〇 ⊗ 〇 ⊗ ▨Deep Vein Thrombosis ⊗ ⊗ ⊗ ⊗ ⊗ 〇 〇 ⊗ ⊗ 〇
Heart attack 〇 ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗
Hemolytic Uremic Syndrome ⊗ ⊗ 〇 〇 U
Kidney Stones 〇 ⊗ ⊗ 〇 ⊗ ⊗ ⊗ 〇 〇 〇 ⊗ ▨Malaria 〇 〇 N N
Meningitis ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ N
Pneumonia (more severe) 〇 ⊗ 〇 〇 ⊗ 〇 ⊗ 〇 〇 〇 ▨Pulmonary Embolism ⊗ 〇 ⊗ ⊗ 〇 〇 〇 ⊗ ⊗ ⊗ ⊗ ▨Rocky Mountain Spotted
Fever⊗
Stroke ⊗ 〇 ⊗ 〇 〇 ⊗ ▨Tetanus ⊗ N
Process couldn't be
started (ex. Too young)
Incorrect
diagnosis
Diagnosis given Triage Advice Given
Symptom Checker
Correct diangosis
listed in top 20
Correct
diagnosis
listed first
Supplemental Table 4: Diagnosis given and triage advice from each symptom checker, stratified by severity of the standardized patient
(SP) vignette. The symptom checkers are organized by the accuracy of listing the correct diagnosis first and the rate of appropriate triage
in descending order from left to right under each respective heading.
Diagnostic Accuracy Rate of appropriate triage
Page 41 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Do
cRe
spo
nse
Fa
mil
y D
oct
or
Isa
be
l
Ask
MD
/
Sh
are
care
Do
cto
r D
iag
no
se
Dru
gs.
com
Sy
mca
t
He
alt
hli
ne
Fre
eM
D
iTri
ag
e
We
bM
D
HM
S F
am
ily
He
alt
h G
uid
e
Ea
rly
Do
c
Sy
mp
tom
ate
Sy
mp
tify
Be
tte
r M
ed
icin
e
Esa
gil
Ma
yo
Cli
nic
ME
Do
cto
r
HM
S F
am
ily
He
alt
h G
uid
e
He
alt
hy
Ch
ild
ren
Ste
ps2
Ca
re
Sy
mp
tify
Sy
mp
tom
ate
Do
cto
r D
iag
no
se
Dru
gs.
com
Fre
eM
D
Fa
mil
y D
oct
or
Ea
rly
Do
c
NH
S
Isa
be
l
Sy
mca
t
He
alt
hw
ise
iTri
ag
e
Symptom Checker
Requires non-emergnt care
Acute otitis media ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ▩Acute pharyngitis ⊗ 〇 ⊗ ⊗ 〇 ⊗ 〇 ⊗ ⊗ ⊗ 〇 ▩ N
Acute pharyngitis ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ N
Acute sinusitis ⊗ 〇 〇 ⊗ 〇 〇 ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ⊗ ▩Back pain with foot drop 〇 ⊗ ⊗ ⊗ ⊗ 〇 ▨Cellulitis ⊗ ⊗ 〇 〇
COPD flare ⊗ ⊗ ⊗ ⊗ 〇 〇 ⊗ ▨ N
Influenza 〇 ⊗ ⊗ 〇 ⊗ 〇 ⊗ 〇 ⊗ ⊗ ⊗ ⊗ ▨Mononucleosis ⊗ 〇 〇 ⊗ ⊗ 〇 ⊗ 〇 〇 ⊗ ⊗ ▩Peptic Ulcer Disease ⊗ ⊗ 〇 〇 ⊗ 〇 ⊗ ⊗ ⊗ ⊗ 〇 ⊗ 〇 ⊗ N
Pneumonia ⊗ 〇 ⊗ ⊗ 〇 ⊗ 〇 〇 ⊗ ▩Salmonella 〇 ⊗ 〇 ⊗ ▩Shingles ⊗ ⊗ ⊗ 〇 ⊗ 〇 〇 〇 ⊗ ⊗ N
Urinary tract infection ⊗ ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ⊗ ⊗ ▩Vertigo ⊗ ⊗ 〇 ⊗ ⊗ N
Diagnostic Accuracy Rate of appropriate triage
Page 42 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review Only
Do
cRe
spo
nse
Fa
mil
y D
oct
or
Isa
be
l
Ask
MD
/
Sh
are
care
Do
cto
r D
iag
no
se
Dru
gs.
com
Sy
mca
t
He
alt
hli
ne
Fre
eM
D
iTri
ag
e
We
bM
D
HM
S F
am
ily
He
alt
h G
uid
e
Ea
rly
Do
c
Sy
mp
tom
ate
Sy
mp
tify
Be
tte
r M
ed
icin
e
Esa
gil
Ma
yo
Cli
nic
ME
Do
cto
r
HM
S F
am
ily
He
alt
h G
uid
e
He
alt
hy
Ch
ild
ren
Ste
ps2
Ca
re
Sy
mp
tify
Sy
mp
tom
ate
Do
cto
r D
iag
no
se
Dru
gs.
com
Fre
eM
D
Fa
mil
y D
oct
or
Ea
rly
Do
c
NH
S
Isa
be
l
Sy
mca
t
He
alt
hw
ise
iTri
ag
e
Symptom Checker
Acute bronchitis ⊗ 〇 〇 〇 〇 ⊗ ⊗ ▧ N
Acute bronchitis 〇 ⊗ ⊗ 〇 ⊗ 〇 ⊗ 〇 〇 〇 ⊗ ⊗ ⊗ N
Acute conjunctivitis ⊗ ⊗ ⊗ ◯ ⊗ ◯ ⊗ ⊗ ⊗ ◯ ⊗ ⊗ ▨Acute pharyngitis ⊗ ⊗ 〇 ⊗ 〇 ⊗ ⊗ 〇 ⊗ ⊗ ▧ N
Allergic rhinitis ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ ⊗ ⊗ 〇 〇 ⊗ ⊗ ⊗ ▧Back pain, unremarkable ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ⊗ 〇 ⊗ ⊗ ▧Bee sting without anaphylaxis ⊗ 〇 ⊗ 〇
Candidal yeast infection ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ 〇 ⊗ ▩Canker sore ⊗ ⊗ 〇 〇 〇 〇 ⊗
Constipation 〇 ⊗ ⊗ ⊗ ⊗ ⊗ ⊗ N U
Eczema ⊗ ⊗ ⊗ 〇 〇 ⊗ 〇 〇 ⊗ ⊗ H
Stye ⊗ 〇 〇 ⊗ 〇 〇 ⊗ 〇 ⊗ ⊗ ▧Viral upper respiratory illness 〇 ◯ ◯ 〇 ⊗ ⊗ 〇 ⊗ 〇 〇 ◯ ⊗ 〇 ⊗ ⊗ ⊗ ▧Viral upper respiratory illness ⊗ 〇 ⊗ ⊗ 〇 〇 ⊗ 〇 〇 ⊗ 〇 ▧Vomiting ⊗ ⊗ 〇 ⊗ 〇 ⊗ 〇 ⊗ 〇 ⊗ ⊗ N N N
Diagnostic Accuracy Rate of appropriate triage
Page 43 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nly
Rate*
Appropriate triage %
(95% CI)
All symptom checkers 301/532 57 (52 to 61)
Without iTriage 287/489 59 (54 to 63)
Without Symcat, Symptomate, and Isabel 249/428 58 (53 to 63)
Without Symcat, Symptomate, Isabel, and iTriage 235/385 61 (56 to 66)
* Number of correct SP evaluations / Applicable SP evaluations. As noted in text, some SP vignettes
could not be applied to a given symptom checker. For example, we could not evaluate an adult SP
vignette if it was a pediatric symptom checker.
Supplemental Table 5: Sensitivity analysis for overall appropriateness of triage advice when
symptom checkers that always provide advice to go to the emergency department are
removed (iTriage) and when symptom checkers that never suggest self-care are removed
(Symcat, Symptomate, and Isabel).
Page 44 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Confidential: For Review O
nlySummary Box
Section 1: “What is already known on this topic”
The public is increasingly using the internet for self-diagnosis and triage advice, and there has
been a proliferation of computerized algorithms called symptom checkers that attempt to
streamline this process. Despite the growth in use of these tools, their clinical performance
has not been thoroughly assessed.
Section 2: “What this study adds”
Our study suggests that symptom checkers have deficits in both diagnosis and triage, and their
triage advice is generally risk-averse.
Page 45 of 45
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960