ORIGINAL ARTICLE

Access to Radiology Reports via an

Online Patient Portal: Experiences of
Referring Physicians and Patients
Dan Henshaw, MD a, Grant Okawa, MD a, Karen Ching, MD a, Terhilda Garrido, MPH b,
Heather Qian, MBAb, Joanne Tsai, MPH b

Abstract
Purpose: Few organizations have reported providing radiology reports to patients via an electronic health record patient portal. The
authors describe the process of manual release of reports made by referring physicians, and patients’ and referring physicians’ experiences
during the first year that release through the portal was available.
Methods: A survey of 508 patients assessed perceived accessibility and importance of portal-released radiology reports, and commu-
nications with referring physicians before and after the release. A survey of 48 referring physicians and a group interview assessed the
utility of releasing reports, preferences regarding automatic release, and workload impact. Data were analyzed using descriptive statistics
and qualitative methods.
Results: A total of 74% (377) of patients found reports easy to access, and 88% (446) reported that the ability to do so was important.
In all, 49% (250) of patients were contacted by their referring physician before report release, and 25% (156) contacted their physician
for more information after viewing a report. Of the referring physicians, 88% (42) found that releasing reports to patients was useful.
Auto-release of x-ray reports, with a 1-week delay, was preferred by 58% (28), but they were more reluctant to auto-release CT and MRI
reports. A total of 86% (41) of referring physicians reported that follow-up emails, telephone calls, and office visits were unchanged or
had decreased.
Conclusions: Referring-physician release of radiology reports via the online portal is important to patients, useful to referring physicians,
and does not affect referring-physician workloads. A delay between reporting results to referring physicians and releasing them to patients
allows time for needed physician-patient communication.
Key Words: Radiology, patient-centered care, electronic health record, qualitative research
J Am Coll Radiol 2015;12:582-586. Copyright
2015 American College of Radiology

INTRODUCTION for which direct reporting of a summary to the patient, in

Providing patients with unfettered access to their own lay terms, is mandated [3].
health information is a pivotal step toward redesigning Patients’ experience waiting for a referring physician
the health care system to better meet their needs [1]. to communicate radiology results is a stressful and
However, access to radiology results has traditionally frustrating experience and the authors perceive that it
been tightly controlled; reports have functioned as pro- contributes to unnecessary delays in receiving needed
prietary communication between radiologists and refer- care [4]. In addition, many patients prefer to obtain
ring physicians, who then communicate results to much more detail about radiology results in a written
patients [2]. A broad exception is mammography results, format than is provided by a typical verbal report in
person or over the telephone [4]. Radiologists and
a
Hawaii Permanente Medical Group, Honolulu, Hawaii. referring physicians both recognize that patients are
b
Health Information Technology & Transformation Analytics, Kaiser dissatisfied with traditional results-reporting methods
Permanente, Oakland, California. and may themselves be dissatisfied with these methods,
Supported by Kaiser Permanente.
at least in part because of the potential for results to
Corresponding author and reprints: Heather Qian, MBA, Health Information
Technology & Transformation Analytics, Kaiser Permanente, 1800 Harrison “fall through the cracks” [5]. However, although they
St, 24th Floor, Oakland, CA 94612; e-mail: [email protected]. believe that patients should have access to personal

ª 2015 American College of Radiology

582 1546-1440/15/$36.00 n http://dx.doi.org/10.1016/j.jacr.2015.01.015
health information, eg, through an online patient portal the results “inbasket” tab in KP HealthConnect by
integrated with an electronic health record (EHR), highlighting a report and clicking on a “kp.org release”
referring physicians and radiologists may be concerned button. In addition, they can enter comments for pa-
that patients will not adequately understand radiology tients to view on the portal along with the radiologist’s
reports, resulting in greater anxiety for patients and report.
increased demands on provider time [5]. The option of releasing radiology reports was enabled
In a 2012 simulation study, 75% of patients for all Kaiser Permanente Hawaii physicians. Although
preferred notification of radiology results by portal for only primary care providers (PCPs) received targeted
normal, indeterminate, and abnormal findings [6]. communications alerting them that this option was
However, few reports exist of organizational experi- available, some specialty physicians released reports as
ence with providing radiology results to patients via an well. In 2013, a total of 133 referring physicians (103
EHR-integrated online patient portal. In 2013, Kaiser PCPs) released a total of 9,311 radiology reports, an
Permanente Hawaii enabled referring physicians to average of 837 per month. X-rays accounted for 5,468
manually release radiology reports to patients via the (59%) of all released reports (Table 1). The proportions
online portal. The purpose of this descriptive and of released reports that were of x-ray and ultrasound
qualitative report is to describe patterns of manual studies (59% and 13%, respectively) were approximately
release of radiology reports by referring physicians, and comparable to the proportion of all studies of those types
the experiences of patients and referring physicians completed in 2013 (61% and 22%, respectively). Of the
during the first year that the option to release radiology 9,311 released reports, 4,643 (50%) were for studies
reports was available. completed before 2013, which referring physicians could
release using a “release all” option; 4,668 (50%) were for
METHODS studies completed in 2013. Referring physicians released
2,474 (53%) of the latter to patients the same day they
Setting became available, and an additional 1,680 (36%) within
Kaiser Permanente Hawaii is 1 of 7 regions of the nation’s a week of availability; the remaining 513 (11%) studies
largest not-for-profit integrated health care delivery sys- were released
2 weeks after availability. Patients viewed
tem. Approximately 230,000 members receive all needed a total of 3,527 (75%) of 4,668 reports created and
care from 500 physicians, representing all clinical spe- released in 2013, and viewed 3,108 (88%) of these
cialties, and 4,500 employees at 1 Kaiser Permanente- within 1 week. Of reports created and released in 2013, a
owned medical center and 18 medical office buildings, on total of 3,394 (73%) were released by PCPs, a volume
3 islands. In 2004, Kaiser Permanente Hawaii imple- equal to 7.3% of PCP-ordered imaging with completed
mented KP HealthConnect, an EHR integrated across all reports.
care settings that includes an online patient portal: kp.org.
As of March 2014, a total of 51% of Internet-using Kaiser
Survey Methods and Data Analysis
Permanente Hawaii members aged
13 years were
In February 2014, we surveyed patients and referring
registered on kp.org.
physicians about their experiences with sending and
receiving radiology reports via the patient portal. The
Release of Radiology Reports online patient survey was distributed via email to
In January 2013, Kaiser Permanente Hawaii began
enabling referring providers to manually release radi-
ology reports to patients, excluding images, via the on- Table 1. All released reports, by imaging modality
line portal. Referring providers include both physicians Number of studies
and a small minority of physician extenders (physician released (%)
assistants and nurse practitioners); for convenience, all Modality N ¼ 9,311
are collectively termed “referring physicians” here. With X-ray 5,468 (59)
the exception of mammograms, for which a process for Ultrasound 1,256 (13)
releasing results already existed, and inpatient and CT 1,011 (11)
MRI and/or MR angiography 793 (8)
emergency department imaging studies, the report pre-
Biodynamic 530 (6)
pared by a radiologist is released to the patient in its Nuclear medicine 253 (3)
entirety. Referring physicians release results from within

Journal of the American College of Radiology 583

Henshaw et al n Radiology Reports
2,450 Kaiser Permanente Hawaii members who had Provider Survey
received reports in this way. An online survey was Forty-eight referring physicians responded within 2
distributed via e-mail to 79 referring physicians, weeks and were included in the data analysis, reflecting a
comprising 73 PCPs and 6 specialty physicians (in response rate of 61%. Their responses were anonymous,
rheumatology, oncology, obstetrics and gynecology, so the proportion of primary care versus specialty phy-
urology, allergy, and nephrology) who had released sicians is unknown. A total of 42 (88%) referring phy-
reports via the patient portal (patient and referring- sicians agreed that releasing radiology reports to patients
physician surveys are available in the Online via the portal was useful; of these, 26 (54% of all re-
Appendix 1). We analyzed survey data using descrip- spondents) strongly agreed that releasing radiology re-
tive statistics. The Kaiser Permanente Hawaii Institu- ports was useful. When asked if they would prefer to
tional Review Board assessed this project as being in have reports automatically released 1 week after they
the category of quality improvement and therefore not were completed (so they could eliminate the time
subject to oversight. required to manually release reports and have an op-
portunity to review the results with the patient), 13
(27%) referring physicians strongly agreed, and 10
RESULTS
(21%) strongly disagreed. A total of 28 (58%) referring
Patient Survey physicians strongly or somewhat agreed to having a
A total of 508 patients completed the survey within 2 preference for auto-release, and a total of 20 (42%)
weeks and were included in the data analysis, reflecting strongly or somewhat disagreed about preferring auto-
a response rate of 21%. Overall, 239 respondents release.
(47%) strongly agreed that they could easily access When asked about the workload impact of releasing
radiology reports on the patient portal, and an addi- radiology reports, 35 (73%) referring physicians re-
tional 138 (27%) somewhat agreed that they found ported that their follow-up work related to radiology
access easy. A total of 446 (88%) members agreed that reports was unchanged. Six (13%) reported decreased
being able to access radiology reports on the patient follow-up work in the form of e-mails, telephone calls,
portal was important to them; of these, 382 (75% of all and office visits from patients; 5 (10%) reported
respondents) agreed strongly. Thirty-eight (7%) re- increased follow-up work; and 2 (4%) lacked a response
spondents strongly disagreed that online access to to this item. A final item assessed reasons that referring
radiology reports was important. physicians stopped manually releasing radiology reports;
A total of 250 (49%) patients reported that their 21 (44%) referring physicians responded that the
physician spoke to them or sent a secure e-mail message question was not applicable because they were still
about radiology results before releasing the report on releasing reports.
the patient portal; 109 (21%) reported no communi- Among those who had stopped releasing reports, the
cation with a physician before gaining access to radi- most frequent (13; 27%) reason was that they had
ology reports. The remaining 149 (30%) respondents forgotten that the option was available. Four (8%)
either did not recall communicating with a physician referring physicians had stopped because of concerns
before a report was released, or lacked a response to this about potential confusion and anxiety among patients; 1
item. When asked if they contacted a physician by e- (2%) had not had the right opportunity to release a
mail, telephone, or an office visit after reviewing a radiology report, and 1 (2%) selected the “other”
radiology report, 286 (56%) patients responded that response option. Eight (17%) referring physicians lacked
they did not require any follow-up with a physician a response to this item. Respondents supplied additional
about the reports, and 126 (25%) reported contacting a comments related to the need to explain results to pa-
physician about the report. Ninety-six (19%) patients tients, and mixed opinions about auto-release of radiology
either did not know if they had contacted a physician reports (Online Table 3).
after reviewing a radiology report, or lacked a response
to this item. Patients offered additional comments
reflecting a positive experience; some indicated a desire DISCUSSION
for increased access, including the ability to view im- Survey results indicated that the practice of manual
ages, and others wanted results to be reported in lay release of radiology reports was well received by the
terms (Online Table 2). majority of patients, and most referring physicians

584 Journal of the American College of Radiology

Volume 12 n Number 6 n June 2015
found it useful. In addition, releasing reports to patients expectation that physicians would supply an explanation
had a negligible impact on referring-physician work- for all released results, even those that were normal. Most
loads. Few physicians stopped releasing radiology re- PCPs at the meeting felt that some communication with
ports to patients via the online portal, out of concern patients should occur when radiology reports were
that doing so would generate confusion and anxiety for released, and that using the automatic messaging function
patients. within KP HealthConnect was the most efficient way to
Strengths of our report include the fact that it is enable this to happen. Smart phrases, which are stan-
among the first to document an organization’s experi- dardized text phrases entered by keystroke shortcuts,
ence with releasing radiology reports to patients. In increased efficiency.
addition, our observation period encompassed an entire Our second question was whether using lay language
calendar year, to capture the impact of a relatively in radiology reports released to patients would prevent the
mature program, and we report on the experiences of need for physician communication to patients at the time
both patients and referring physicians. Limitations to of release. PCPs agreed that they still needed to
our report include the fact that it took place within an communicate results to patients, but a standardized
integrated health care delivery system; the generaliz- format for the report itself would reduce the amount of
ability of our results is unknown, although we have no required explanation. They were not in favor of moving
reason to believe that Kaiser Permanente Hawaii’s to lay language, because reports would then no longer
organizational or membership characteristics influenced serve their needs. This finding encapsulates the tension
the findings. inherent in making available to patients a report that is
A similar release of radiology reports can take place in intended to facilitate communication between pro-
any setting that has an online patient portal. Another fessionals and provide comprehensive documentation.
limitation is that survey participants were self-selected; Suggestions for improving the clarity of radiology reports,
respondents might have differed from nonrespondents including reducing the use of jargon and uncommon
in ways that affected their experiences and responses. abbreviations, have been made [2]. In the presence of
Finally, we note that the overall volume of current radi- more-complicated and detailed reports, such as those for
ology reports that were released to patients is very low; CT and MRI scans, communication at the time of release
different findings might result from a substantially higher could be limited to the most important information—the
volume of released reports. bottom line—as opposed to encompassing incidental
The survey data suggest that some arguments against findings. However, we note, as have others, that
releasing radiology reports to patients are unfounded [7]. improving the quality of radiology reports to facilitate
Less than 10% of referring physicians were concerned transparency with patients represents a substantial un-
that releasing reports generated confusion and anxiety dertaking for radiologists [2].
among patients, and workload increases related to follow- Our third question probed for underreported
up of radiology results were infrequently reported. workload increases related to follow-up for radiology
Referring physicians selected reports for release after reports. Responses from PCPs at the meeting rein-
reviewing them, which prevented patients from having forced the survey findings; workload increases were
access to reports before physician review and provided a insignificant, even though some providers reported an
“filtering” mechanism in terms of which reports were increase in telephone calls related to incidental find-
released [7]. ings. The final question pertained to auto-release of x-
However, the survey also provided data indicating ray reports. PCPs reported no strong opposition, but
the areas we needed to address. Patients noted that the they expressed disparate opinions about the optimum
language of radiology reports was not always intelligible, timing for auto-release. Some were in favor of releasing
and less than half received communication from a within a short period of time because patients want and
referring physician about reports that were accessible expect to get their results soon after examination
online. Strong physician preferences for and against completion; others were concerned about being unable
auto-release of reports indicated that more information to reach patients by telephone to review the report
was needed. before its release. PCPs were particularly concerned
Consequently, we held a discussion at a primary about being unable to reach their colleagues’ patients
care-department meeting to solicit input on 4 questions. when they were providing coverage for an extended
First, we wanted to know if it should be an organizational period of time.

Journal of the American College of Radiology 585

Henshaw et al n Radiology Reports
 In general, PCPs expressed more reluctance about auto- n Many referring physicians favored auto-release of
release of CT and MRI reports, compared with x-ray re-
simpler types of reports, such as x-ray, with an
ports, for 2 reasons. First, the increased likelihood of a
automatic delay to allow them time to communi-
clinically significant and worrisome finding makes it even
cate with patients about results.
more important that patients and physicians discuss the
results before the report is released. Second, these reports
are generally longer and more complex than are x-ray re- ADDITIONAL RESOURCES
ports; PCPs were concerned that auto-release would Additional resources can be found online at: http://dx.
generate telephone calls or e-mails from patients who did doi.org/10.1016/j.jacr.2015.01.015.
not understand their reports.
In summary, manual release by referring physicians of REFERENCES
radiology reports via an online patient portal is important 1. Institute of Medicine. Crossing the quality chasm: A new health system
for the 21st century. Washington, DC: National Academies Press; 2001.
to patients and useful to physicians and has minimal 2. Bruno MA, Petscavage-Thomas JM, Mohr MJ, Bell SK, Brown SD.
impact on referring-physician workloads. Auto-releasing The “open letter”: radiologists’ reports in the era of patient web portals.
reports may work best for x-rays, and a delay between J Am Coll Radiol 2014;11:863-7.
3. US Congress. Mammography Quality Standards Act of 1992, pub l
reporting results to physicians and releasing them to pa- no 102-539, as amended by Mammography Quality Standards
tients would allow needed communication before patients Reauthorization Act of 1998 and 2004, pub l no 105-248, 42 usc §
access their radiology reports. 263(b). Available at http://www.fda.gov/Radiation-EmittingProducts/
MammographyQualityStandardsActandProgram/Regulations/ucm110823.
htm.
4. Johnson AJ, Easterling D, Williams LS, Glover S, Frankel RM. Insight
TAKE-HOME POINTS from patients for radiologists: improving our reporting systems. J Am
Coll Radiol 2009;6:786-94.
n Release of radiology reports via the online patient 5. Johnson AJ, Frankel RM, Williams LS, Glover S, Easterling D. Patient
portal of an EHR is important to patients, and access to radiology reports: What do physicians think? J Am Coll Radiol
useful to referring physicians. 2010;7:281-9.
6. Johnson AJ, Easterling D, Nelson R, Chen MY, Frankel RM. Access to
n Report release did not significantly change workloads radiologic reports via a patient portal: clinical simulations to investigate
for physicians in the form of telephone calls, e-mails, patient preferences. J Am Coll Radiol 2012;9:256-63.
or office visits with patients who viewed a report. 7. Johnson AJ. Reporting radiology results to patients: keeping them calm
versus keeping them under control. Imaging Med 2010;2:477-82.

586 Journal of the American College of Radiology

Volume 12 n Number 6 n June 2015
APPENDIX 1: PATIENT AND PHYSICIAN Table 2. Patient comments about accessing radiology
SURVEYS reports online
Patient Survey
1. The radiology reports released to me on the portal are It was a positive experience
“My doctor placed a message in my e-mail that was attached
easy to access.
to the report, which is great.”
Strongly agree/Agree/Disagree/Strongly disagree/Missing
“It’s helpful and much more efficient to see it online.”
2. Being able to see my radiology reports on the portal is “I was glad to get my results so quickly and easily.”
important to me. “I am very pleased with the radiology and blood test reports.”
Strongly agree/Agree/Disagree/Strongly disagree/Missing They wanted access to more reports, and to images
3. My doctor spoke to me or sent me a note about the “My biggest issue is that some reports have never been
results of my radiology report before s/he released it to released to me while others have.”
me. “I cannot find all my x-ray results. I assume there are certain
ones that do not get released to patients.”
Yes/No/Don’t know/Missing
“I have had an elbow x-ray, chest x-ray, CT scan, etc. Only
4. After reviewing the radiology report on the portal, I
report showing is for elbow—very insufficient reporting on
contacted my doctor (by email, phone, or office visit) those tests.”
to discuss it. “I would have liked to see the images.”
Yes/No/Don’t know/Missing They wanted a more comprehensible report
“I wish there was more of an explanation or summary as to
Referring Physician Survey what the results mean.”
1. Being able to release radiology reports to my patients “Medical terms not easy to comprehend. It would be better to
via the portal is useful. describe results in layman’s terms.”
Strongly agree/Agree/Disagree/Strongly disagree “Explanation needs to be better for a lay person.”
2. I would prefer to have the reports released automati-
cally 7 days after they are completed. This would give
me enough time to review the results with the patient
while saving me the time needed to manually release Table 3. Physician comments about releasing radiology
each report. reports to patients via an online portal
Strongly agree/Agree/Disagree/Strongly disagree
3. Related to the radiology report being released to my It is helpful to explain results, so releasing reports is
patients, I’ve experienced: beneficial to patients and physicians
About the same follow-up work (phone calls, emails, “Making sure that the reports are easy to read, in layman’s
terms as much as possible, could decrease the amount of
office visits)/Additional follow-up work (phone calls,
call backs and patient concerns. It’s helpful and much more
emails, office visits)/Less follow-up work (phone calls, efficient to see it online.”
emails, office visits)/Missing “Many radiology reports have benign but abnormal findings
4. I have stopped using this feature because: mentioned. I recommend we have smart phrases
Not applicable—I am releasing reports/I forgot this (standardized text entries) to help explain common findings
was available/I am concerned that it might cause that are clinically benign but possibly alarming to the
confusion or anxiety for my patients/I have not had the patient. This will help educate clinicians as well.”
right opportunity/Other/Missing “When I press the kp.org release button from the results tab,
the results are released. I usually (add) some explanatory
comments.”
“I have been attaching a short note with the released results. I
think that this really reduces the follow up email and calls.”
They had mixed feelings about auto-release of radiology
reports
“Let’s move to auto-release.”
“I think that an automatic release could be problematic if you
are trying to reach a patient and playing phone tag during
the week so the patient may see a concerning report
without having spoken with a provider first.”
“Would only release normal results to patients”

Journal of the American College of Radiology 586.e1

Henshaw et al n Radiology Reports