ArticlePDF Available

PACS and Unread Images 1

Authors:
PACS and Unread Images
1
Robert W. Evers, BSRT, David M. Yousem, MD, Tom Deluca, MCSE, Norman J. Beauchamp Jr, MD, Sidney Smith, RTR
In many circles, the justification for purchasing a picture
archiving and communication system (PACS) has already
been settled because of the proliferation of image data. A
PACS is used to acquire medical images digitally from
the various modalities, such as computed tomography
(CT), magnetic resonance (MR) imaging, ultrasound, nu-
clear medicine, and digital radiography (1). The need to
manage, transfer, and transport thousands of images effec-
tively has led to medical and legal arguments for storing
such data electronically rather than producing each image
on film. Moreover, with the dissemination of imaging
centers—as well as one’s referral base, which may extend
beyond state or country boundaries—a PACS becomes an
integral part of running an efficient department. The
transfer of images to all-night reading stations for subse-
quent consolidation of emergency room coverage and the
centralization of reading areas within a hospital or an en-
terprise-wide multisite radiology group, has also made life
without PACS nearly impossible. Without the improve-
ments in speed and bandwidth for the data transfer, how-
ever, these consolidations would be severely limited.
Nonetheless, at a local level, one must still justify the
huge investment in capital for a PACS to the financial
planners of a hospital or outpatient group practice. We
are facing an ever-challenging situation of dwindling bud-
gets, increasing cost pressure, and growing demands to
increase the efficiency and quality of related services (2).
Typically, reductions in film cost and film library person-
nel are cited as a means of “financing” the PACS. PACS
eliminates the film-associated workload, which includes
processing, filing, and manual retrieval of previous studies
from film storage. These steps constitute a considerable
part of the total radiology turnaround time (3). The eco-
nomics of PACS are characterized by higher fixed costs
for the digital infrastructure and lower variable or mar-
ginal costs related to savings in film and personnel (1).
Yet another source of savings in converting to a PACS
is the greater capture of studies obtained within the de-
partment when film is not relied on as the primary means
of viewing images. Heretofore, a common complaint in
radiology departments was that clinicians would remove
film images from the department before the radiologist
could interpret them, rendering these examinations “non-
billable” without the cost of refilming. This is a common
occurrence in most emergency room radiography depart-
ments, where images may be sequestered by the treating
physician before being accessioned for billing or account-
ing. Not only do patients lose an opportunity for expert
radiologic interpretation, but these lost films typically will
not be available for future comparative studies (4). Ac-
countability for them is particularly problematic without
an integrated radiology information system (RIS). The
number of lost studies will theoretically be reduced after
integration of the RIS and PACS. When the radiology
team and referring clinicians have simultaneous access to
all patients’ digital data, patient care will improve. There-
fore, in addition to decreasing the number of unread im-
ages, the use of PACS should help decrease the overall
report turnaround time, permitting referring clinicians to
make swift decisions on treatment options and health care
delivery (4).
At the Johns Hopkins Hospital, Baltimore, Md, the
radiology department elected to implement the PACS in a
piecemeal fashion because of funding considerations. For
a time, the outpatient MR imaging service used a PACS
and soft-copy reading to transfer and interpret images,
while the inpatient MR imaging service continued to use
hard-copy film reading. We were therefore able to com-
pare the rates at which studies were lost with a PACS and
remote reading (the outpatient service) and a film-based
Acad Radiol 2002; 9:1326 –1330
1
From the Russell H. Morgan Department of Radiology and Radiological
Sciences, Division of Neuroradiology, the Johns Hopkins Medical Institu-
tions, 600 N Wolfe St, Phipps B-112, Baltimore, MD 21287. Received July
21; accepted July 22. Address correspondence to D.M.Y.
©
AUR, 2002
1326
Point-Counterpoint
(inpatient) service. We had the unique opportunity to
study the rates of unread studies within the two settings
for the years before (1998) and after (2001) the PACS
was implemented. We hypothesized that conversion to the
PACS and soft-copy reading would lead to a severalfold
reduction in the rate of unread studies.
MATERIALS AND METHODS
During late 2000, the neuroradiology division of our
institution elected to transfer as soft copy the images re-
sulting from all outpatient MR imaging studies, from our
outpatient center to an inpatient central reading area. For
2 years previously CT scans from six sites in the central
reading area had been read on a PACS with soft-copy
interpretation of images. To increase divisional efficiency,
the outpatient MR imaging service was integrated into
this reading arrangement.
Two MR imagers are employed in the outpatient cen-
ter: a GE Medical Systems Signa imager (Milwaukee,
Wis) and a Siemens Medical Systems Vision imager (Er-
langen, Germany). They are used from 7:30 AM to 5:30
PM each weekday but not on weekends. The service is
100% outpatient and more than 95% neuroradiology
based. Studies are transferred through a hospital intranet
and coordinated with a Siemens Magic View workstation.
Studies are read by one neuroradiologist who covers both
the outpatient and inpatient services for that day.
The inpatient MR imaging service uses three GE Signa
imagers. The hours of operation are from 6:30 AM to
11:30 PM for two imagers and from 6:30 AM to 5:00 PM
for the third. The practice is a combination of neuroradi-
ology and body MR imaging, at a ratio of approximately
2:1.
The inpatient studies are read at a site approximately
100 yards from the outpatient MR imaging PACS work-
stations. Functional MR and perfusion MR imaging stud-
ies are obtained nearly exclusively on the inpatient imag-
ers, but the neuroradiologic procedures are otherwise
identical between inpatient and outpatient sides.
After each MR examination is completed, the technol-
ogist completes the necessary paperwork and computer
work within the RIS. At Johns Hopkins, this step is re-
ferred to as “processing a header” and “accessioning a
case.” This header or data entry communicates the fol-
lowing information: (a) exact procedure(s) completed;
(b) appropriate ICD-9 (International Classification of Dis-
eases, Ninth Revision) coding for the procedure(s); (c)
examination times (beginning and ending); (d) all technol-
ogists involved with the case; (e) the section of the MR
imaging department where the examination was done;
(f) the appropriate inpatient admission or outpatient regis-
tration; (g) the ordering attending physician and/or refer-
ring clinician; (h) a unique identifier (accession number)
associated with each procedure (if an order was entered in
the system before the procedure, the accession identifier
is pulled forward to maintain the link to the study); and
(i) a header accession number, created to group all proce-
dure codes within the study for interpretation. Once a
header has been processed, the examination is ready for
interpretation.
Every Friday throughout the year, the department takes
a snapshot of the number of unread studies in each sub-
section of the departmental services. In addition, a read-
out of the unread studies with the patients’ identities,
based on an interface with the RIS, is printed twice daily
for both inpatient and outpatient services and is used by
the film librarians, the physicians, and the chief technolo-
gists to track and regenerate studies as needed. After 30
days, an imaging study that goes unreported is removed
from the unread header list on both of the daily printouts
and the Friday reporting number and is presumed to be
nonbillable. A case in which studies are unread can there-
fore appear on a list for a maximum of 4–5 weeks.
In 2001, the lists of unread studies for each Friday for
both the outpatient MR and the inpatient MR imaging
services were available for 47 weeks. The 5 missing
weeks were randomly dispersed throughout the year and
in part accountable for by vacation times or computer
downtime. We compared the values for the inpatient and
outpatient MR imaging services between 1998 (the year
before PACS was begun in our department) and 2001 (the
year after PACS was integrated into outpatient MR imag-
ing) to determine the effect of the PACS reading of the
outpatient studies. The 1998 data, used as a control, were
collected in the same way as the 2001 data, with 40
weeks of data available for 1998. The mean, median, and
mode with standard deviation (SD) for unread studies
were obtained for inpatient hard-copy and outpatient soft-
copy MR services, and the data for each year from the
two sites were compared by means of a two-tailed paired
t test.
As stated previously, an unread study rotates off the
Friday unread header list if it is 30 days old. In this way,
we ensured that the same studies could not contribute to
more than 4–5 weeks’ values and thereby skew the data.
Additionally, as stated above, the printout of unread head-
Academic Radiology, Vol 9, No 11, November 2002
PACS AND UNREAD IMAGES
1327
ers was used by film librarians each day to retrieve such
studies either electronically or as film to reduce the list.
RESULTS
The mean number of unread studies each Friday for
the soft-copy outpatient service in 2001 was 3.75 (SD,
3.88; median, 3; mode, 1; range, 0–16). In 1998 the mean
hard-copy outpatient value was 2.33 (SD, 2.45; median,
2; mode, 0; range, 0–10). This reflects a 62% increase in
the number of unread headers after PACS implementa-
tion. The mean number of unread studies each week for
the hard-copy inpatient service was 21.53 in 2001 (SD,
9.36; median, 20; mode, 14; range, 10–53), compared
with 28.32 in 1998 (SD, 14.26; median, 23; mode, 15;
range, 12–78), reflecting a 24% decline.
A t test was performed to compare the differences
from year to year and location to location. The difference
in the number of unread studies between inpatient and
outpatient services was statistically significant in both
1998 and 2001 (P .001 for both years), with more un-
read headers on the inpatient side. There was a statisti-
cally significant drop in unread studies for the inpatient
service between 1998 and 2001 (P .014). On the out-
patient side, there was an increase approaching statistical
significance (P .067) in the number of unread studies
between 1998 and 2001. The difference in change be-
tween the two sites was significant (P .01).
DISCUSSION
The integration of the PACS with the RIS addresses
the issue of the missing or misplaced studies that would
otherwise go unread. Previously, the retrieval of lost films
was completed manually, and it was very time consuming
because one had to search the hospital for lost films or
electronically reload and refilm missing studies for inter-
pretation. Studies comparing retrieval times for PACS and
film-based systems have shown substantial reductions in
retrieval times for the digital method, often down from
several hours to mere seconds (5).
The results of our study demonstrated that the rate of
unread studies was higher the year after the PACS was
implemented in the outpatient MR imaging service than it
was the year before implementation. In the same period,
however, there was a reduction in unread headers for the
hard-copy film-based system in the inpatient service. This
was opposite to what we had predicted. While several
factors could have biased our study, the discrepancy in
the mean outpatient values (3.75 cases per week, up from
2.33) shows an unsettling trend.
Several factors were controlled for in this study. In
both the inpatient and the outpatient settings, the MR im-
aging technologists entered the patient information into
the RIS after completion of the MR imaging examination.
The capture of the studies is therefore believed to be
complete. Physicians are unable to report cases unless the
patient information has been entered into the RIS, and the
accession number must be included in all case-related
dictation. Also, both the inpatient and the outpatient ser-
vices had daily reporting of unread header lists that were
used by film librarians to retrieve cases for 1998 and
2001. The outpatient and inpatient MR imaging technolo-
gists work as a unit with cross-coverage, and they both
report to the central chief technologist, who administrates
both areas. The same neuroradiologist who was on the
outpatient service covered the inpatient service each day.
The same reporting mechanism for the Friday unread
header list count was used, at the same time of day, for
both the inpatient and outpatient services for the same
years.
There were some differences between inpatient and
outpatient settings, however, that could have contributed
to the greater absolute number of unread headers in the
inpatient service. For example, referring clinicians were
more likely to remove inpatient acute studies from the
reading board than outpatient studies. In addition, the in-
patient studies included more sophisticated examinations,
such as perfusion and functional MR imaging studies,
which might be more likely to be lost in the daily work-
load. The number of images per patient examination was
therefore expected to be greater on the inpatient side than
on the outpatient side, where perfusion imaging, MR an-
giographic studies, and functional MR imaging examina-
tions were less common. Another difference is that body
MR imaging examinations were an important part of the
volume on the inpatient side, while that on the outpatient
side was entirely neuroradiologic. By comparing the same
services for 2 years, 1998 and 2001, we hoped to address
these issues.
Other differences may bias the results. The film librari-
ans for the inpatient MR imaging service were different
from those for the outpatient service. The MR imaging
technologists assigned to the outpatient service in general
were different from those on the inpatient service, al-
though cross-coverage did occur. For the outpatient MR
imaging service, a voice recognition system was used for
dictating reports, whereas the inpatient service used medi-
EVERS ET AL
Academic Radiology, Vol 9, No 11, November 2002
1328
cal transcriptionists. On the outpatient side the reports
were signed off immediately after review by the attending
radiologist using voice-recognition dictation, while on the
inpatient side the interval from dictation to transcription
averaged 12 hours and the delay from transcription to
verification of the study averaged 6 hours. Again, these
delays due to the transcription service on the inpatient
side were hypothesized to favor a greater number of un-
read studies for hard-copy reading and do not explain the
trends within each setting between 1998 and 2001.
What are the reasons for the observed increase in un-
read headers with use of a PACS? Why were there more
unread studies in 2001 than in 1998? Some factors point
to the importance of having a fully operational PACS
integrated with the RIS. Both are still in development in
our department as the PACS-RIS system is constructed
division by division at our institution. Because of budget-
ary constraints, a department-wide PACS-RIS solution
could not be funded, even though the costs of storage and
infrastructure have decreased enough that the cost-effec-
tiveness of the technology can be demonstrated positively
for some larger institutions (1).
Unread headers could certainly be created by break-
downs in the numerous steps integrating the system. A
study may be entered by a technologist but not transferred
to the viewing workstation or the archive. It may be sent
to the workstation or archive but have transmission errors
resulting in its lack of receipt at the remote location. A
study may be sent to the incorrect location, although
many automatic routing schemes based on imager work
lists are employed to prevent such occurrences. A study
may be received at the imaging workstation but ignored.
As studies are reviewed and reports are dictated by the
attending radiologist, they are marked as “signed off,”
which eliminates them from the unread list on the work-
station. A study may inadvertently be marked this way
even though it has not been read yet. A report of the
study may be transcribed with the voice-recognition dicta-
tion system but then not merged into the RIS, and there-
fore the study might still be listed as unread. Since the
attending physician has electronically signed the report
with the voice recognition system, he or she would as-
sume that the study has been reported and would mark it
at the workstation as signed off, thus unintentionally cre-
ating an unread header entry. Duplicate accessioning of
studies may also occur if case studies, Current Proce-
dural Terminology codes, or ICD-9 codes are revised af-
ter the fact and the initial study header is canceled or
changed. The report dictated with voice recognition
would be on a canceled or incorrectly headered entry, and
the attending radiologist would unknowingly mark the
case as signed off.
On the inpatient side, studies could appear on the un-
read header list if the images were removed by the clini-
cians, if the images were not hung on viewing boards or
were mistakenly filed by the film librarians, if the attend-
ing physicians failed to read the studies, if the technolo-
gists failed to film the studies, or if there was an error in
transcription. There are many human contacts and checks
and balances with hard-copy reading, providing opportu-
nities for someone (technologist, film librarian, physician,
transcriptionist, or administrator) to discover the unread
study. With a PACS-RIS solution, however, there are few
human interfaces to point out the discrepancies between
studies obtained and studies read.
It was impossible to review the unread header list ret-
rospectively to determine the specific causes of unread
studies from 1998 and 2001, since they were more than
30 days old when this report and survey were initialized.
Nonetheless, a prospective analysis of 20 unread studies
from 1 month in 2002 on the outpatient side showed that
50% of the unread header studies were due to lack of
receipt at the imaging workstation caused by a “network
error” or failure of the technologist sending the study.
Thirty percent were due to physicians reporting and sign-
ing off on studies that then never reached the RIS, mostly
because of incorrect headers (including incorrect acces-
sion numbers) or snafus in the voice recognition system.
Ten percent of the unread studies had been marked as
signed off even though no dictation had been performed.
The remaining 10% represented errors in order entry, in
which the wrong study, patient number, or dates were
entered into the header information and the reported study
on that patient had not really been performed.
In a perfect world, one would assume that a PACS-
RIS solution would result in fewer cases unavailable for
follow-up, as we had hypothesized at the inception of this
study. It is highly likely that the number of MR imaging
records lost in our department represents just a fraction of
the total number of MR imaging examinations that re-
mained unbilled and/or incomplete. Our study assumed
that the entry of all examination orders was complete
100% of the time. However, during the course of an ac-
tual workday, it is not unusual for cases not to be acces-
sioned by the technologist or clerical staff. Orders lacking
headers would not be detected by the accounting system
in which unread headers are tallied, since they would
never have had headers entered. Such occurrences may
Academic Radiology, Vol 9, No 11, November 2002
PACS AND UNREAD IMAGES
1329
have skewed our data in favor of the inpatient imaging
center, where chances are greater that a filmed study
might be removed from the department before its header
is entered. As one integrates order entry into the schedul-
ing system, however, such occurrences are less likely.
The practical implications of our findings can be di-
vided into information technology suggestions and per-
sonnel suggestions. Our information technology recom-
mendations are as follows:
1. Piecemeal integration of PACS and the RIS may
lead to inefficiencies that impair the system’s tracking
mechanisms. Soft-copy reading without an integrated RIS
may result in financial losses rather than the expected
gains, if one does not account for the potential loss of
studies in the system.
2. If headers are canceled or revised, the reports dic-
tated under those accession codes will not be transferred
automatically without explicit instructions and work or-
ders coordinating the PACS and the RIS. This problem
should be addressed.
3. Network glitches or downtime may result in insuffi-
cient data transfer. Backups must be allocated.
4. Automated transfer of studies through the end-ex-
amination command is preferable to reliance on personnel
to send images separately and manually through the net-
work.
5. As the daily queries for unread headers are con-
structed, the RIS should automatically send those unread
studies back to the workstations for reporting.
6. Daily analysis of unread headers is recommended to
identify systemic errors.
The PACS and soft-copy reading are expected to liber-
ate personnel who would otherwise be filming studies,
filing film images, and transporting studies from various
locations. Our recommendations regarding better utiliza-
tion of personnel include the following:
1. Some percentage of a single full-time equivalent
should be devoted to discovering, retrieving, and account-
ing for any unread headers each day. If unread studies are
reduced by 5 per day on a daily volume of 100 MR im-
aging studies (typical of our department), the savings will
support one salaried employee.
2. An employee dedicated to transcription or voice
recognition errors should be assigned to coordinate cases
in which there have been changed or deleted headers,
changed names (female X in the ER to “Jane Doe”),
and/or errors in sign-off.
3. If automation of transfer of images cannot be per-
formed on end-examination commands, then it is criti-
cally important to instruct technologists on the timely and
uniform network commands to transfer images. They
must be able to confirm the receipt and not just the send-
ing of studies.
4. If gaps in the study completions are identified, then
each site of imaging should manually fax a list of com-
pleted studies at the end of the day to someone in the
department who can verify they were delivered to the
right location and appropriately reviewed.
In conclusion, our results suggest that when one is try-
ing to convince an institution’s financial officers of the
advantages of purchasing a PACS, one should focus on
the improved efficiency of technologists, physicians, and
staff; reduced expenditures on film and personnel; and the
ability to account completely for studies within the sys-
tem. The costs associated with full PACS-RIS integration
have been steadily decreasing. Because of continuing cost
decreases for all forms of digital technology, this trend
will continue to enable cost-effective solutions for many
more hospitals and clinics (1). However, even in a well-
run and integrated PACS-RIS enterprise, the system may
break down at numerous points, resulting in unread stud-
ies. This problem seems to be amplified if the PACS and
the RIS are funded and implemented piece by piece with
no fully integrated tracking system to identify and flag
potential pitfalls or losses. A rigid system of surveillance
for unread studies during the turmoil of daily department
functioning is critical. This is an argument for full inte-
gration of a PACS-RIS tracking system even though the
initial start-up costs may seem prohibitive.
REFERENCES
1. Arenson R, Andriole KP, Avrin DE, Gould RG. Computers in imaging
and health care: now and in the future. J Digit Imaging 2000; 13:145–
156.
2. Wendler T. Workflow management: a perspective for more efficient ra-
diology. Appl Radiol Online 1999; 28:1.
3. Twair A, Torreggiani W, Mahmud SM, Ramesh N, Hogan B. Significant
savings in radiologic report turnaround time after implementation of a
complete picture archiving and communication system (PACS). J Digit
Imaging 2000; 13:175–177.
4. Huda W, Szeverenyi N. The filmless radiology department: a primer.
Appl Radiol Online 1999; 28:1–7.
5. Andriole KP. Archiving: a bigger, better box? Decis Imaging Informatics
2001; 14:7–10.
EVERS ET AL
Academic Radiology, Vol 9, No 11, November 2002
1330
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Early picture archiving and communication systems (PACS) were characterized by the use of very expensive hardware devices, cumbersome display stations, duplication of database content, lack of interfaces to other clinical information systems, and immaturity in their understanding of the folder manager concepts and workflow reengineering. They were implemented historically at large academic medical centers by biomedical engineers and imaging informaticists. PACS were nonstandard, home-grown projects with mixed clinical acceptance. However, they clearly showed the great potential for PACS and filmless medical imaging. Filmless radiology is a reality today. The advent of efficient softcopy display of images provides a means for dealing with the ever-increasing number of studies and number of images per study. Computer power has increased, and archival storage cost has decreased to the extent that the economics of PACS is justifiable with respect to film. Network bandwidths have increased to allow large studies of many megabytes to arrive at display stations within seconds of examination completion. PACS vendors have recognized the need for efficient workflow and have built systems with intelligence in the mangement of patient data. Close integration with the hospital information system (HIS)-radiology information system (RIS) is critical for system functionality. Successful implementation of PACS requires integration or interoperation with hospital and radiology information systems. Besides the economic advantages, secure rapid access to all clinical information on patients, including imaging studies, anytime and anywhere, enhances the quality of patient care, although it is difficult to quantify. Medical image management systems are maturing, providing access outside of the radiology department to images and clinical information throughout the hospital or the enterprise via the Internet. Small and medium-sized community hospitals, private practices, and outpatient centers in rural areas will begin realizing the benefits of PACS already realized by the large tertiary care academic medical centers and research institutions. Hand-held devices and the Worldwide Web are going to change the way people communicate and do business. The impact on health care will be huge, including radiology. Computer-aided diagnosis, decision support tools, virtual imaging, and guidance systems will transform our practice as value-added applications utilizing the technologies pushed by PACS development efforts. Outcomes data and the electronic medical record (EMR) will drive our interactions with referring physicians and we expect the radiologist to become the informaticist, a new version of the medical management consultant.
Article
One of the important advantages of the picture archiving and communication system (PACS) is the time saved in comparison with the conventional system. A group of 100 radiologic studies done in a conventional radiology department is compared with another group of the same number done in a completely filmless PACS department to assess the difference in the radiologist report turnaround time. There was a statistically significant (P<.00001) decrease in the median imaging-to-dictation time (IDT) of the PACS group (3 hours and 40 minutes) in comparison with the pre-PACS group (25 hours and 19 minutes). This can be attributed to the fact that PACS eliminates all the workload associated with hard copy films, thus, improving the department’s efficiency and decreasing the number of lost films.
Archiving: a bigger, better box?
  • Andriole
Andriole KP. Archiving: a bigger, better box? Decis Imaging Informatics 2001; 14:7–10. EVERS ET AL Academic Radiology, Vol 9, No 11, November 2002 1330
The filmless radiology department: a primer
  • W Huda
  • N Szeverenyi
Huda W, Szeverenyi N. The filmless radiology department: a primer. Appl Radiol Online 1999; 28:1-7.
  • Evers Et
EVERS ET AL Academic Radiology, Vol 9, No 11, November 2002