The duty-hour standards mandated on July 1, 2003, by the Accreditation
Council for Graduate Medical Education (ACGME) generated a great deal of
controversy regarding resident education. Proponents of the guidelines cite
the potential decrease in medical errors due to resident fatigue and the
improved education, academic performance, and well-being of the residents.
Detractors believe that increasing the number of shift changes results in a
lack of continuity of care that may lead to an increase in medical errors and
a decrease in operative and clinical
experience1-3.
A recent survey of orthopaedic residents by Zuckerman et al. found that the
new work rules had improved the residents' quality of life, allowing them more
rest, more time for reading, and more time to spend with their
families4. However,
the same survey demonstrated that the residents felt their operative
experience had been negatively impacted. A number of concerns have arisen, and
they include the possibility of decreased operative experience for residents
in surgical
specialties5.
Weatherby et al. recently presented research showing that the number of cases
performed by postgraduate year (PGY)-2 and PGY-3 residents had
decreased6. This
finding led to some discussion as to whether the length of residency training
programs should be
increased2.
In order to investigate the effect of the eighty-hour workweek on the
operative experience of residents, we decided to review and compare the number
of procedures performed by orthopaedic residents in our program before and
after the institution of the eighty-hour workweek guidelines.
A retrospective study of the operative experience of residents was
undertaken at our academic institution after institutional review board
exemption was obtained. The study included all residents in training at our
residency program between July 1, 2001, and June 30, 2005. There were a total
of nineteen attending surgeons at eight different institutions involved in
resident education during the study period. Eleven of the nineteen attending
surgeons were considered to be full-time academic faculty. Four of the eight
institutions were considered to be teaching hospitals. During the study,
residents spent a minimum of ten weeks each academic year on private-practice
rotations.
The total number of Current Procedural Terminology (CPT) codes per resident
(PGY-2 through PGY-5) logged from July 1, 2001, to June 30, 2005, were
collected. The number of CPT codes per resident was acquired from the ACGME
database under the search heading "resident experience." The data
were split into two groups on the basis of the codes logged per resident per
year (resident work year). Group I (thirty-nine resident work years) included
CPT codes logged for the two years prior to the implementation of the
eighty-hour workweek (July 1, 2003), while Group II (forty resident work
years) included the codes logged for the following two years. The average
number of CPT codes logged per resident per year was determined.
The total number of CPT codes in each group was then divided by year of
training into four postgraduate-year subgroups, PGY-2, PGY-3, PGY-4 and PGY-5.
The average number of CPT codes for each year of training was determined. The
four groups were then combined into junior residents (PGY-2 and PGY-3) and
senior residents (PGY-4 and PGY-5), and again the average number of CPT codes
was determined for each respective group. All groups were then compared before
and after July 1, 2003.
Statistical Methods
The independent t test was used to compare the numbers before the
eighty-hour workweek (Group I) and after the eighty-hour workweek (Group II).
A p value of <0.05 was selected as significant for this study. Microsoft
Office Excel software (version 2003; Microsoft, Redmond, Washington) was used
for statistical analysis.
The present study indicates that, in our program, the implementation of the
ACGME duty-hour standards had not decreased the number of operative CPT codes
logged per resident. In fact, there was an overall slight increase in the
number of codes logged per resident per year, and thus the work-hour
restrictions had not quantitatively affected the operative experience of
residents. Our findings are similar to those of Spencer and Teitelbaum, who
reviewed the work hours of surgical residents on a subspecialty service for
the year prior to and the year following the institution of the eighty-hour
work-week restrictions and found no change in the residents' operative
experience7. A
survey of surgery residents by Vetto and Robbins, which was related to the
impact of the eighty-hour workweek and specifically to cancer education and
operative experience, found that the majority of the respondents felt there
was no change in their
experience8.
McElearney et al. retrospectively compiled case-log data and operating-room
records of surgery residents for 2002 and 2003 and found no difference in the
number of operative cases in a comparison of residents at all postoperative
years9. The
Residency Review Committee-Surgery reviewed surgical (volume) experience for
general surgical residents before and one year after the implementation of the
ACGME duty-hour restrictions and found no significant change in the overall
surgical experience for major procedures per
resident10.
Our findings are in contrast to those of Weatherby et al., who found a 21%
decrease in the number of operations performed by PGY-2 and PGY-3 orthopaedic
residents following the institution of the eighty-hour workweek
regulations6. We
found no significant change in the number of procedures performed by residents
at any level of training. Although not significant, our findings showed a
slight increase in the number of procedures performed by residents at the
PGY-2, PGY-3, and PGY-5 levels (mean increase, 52.1, 47.6, and 106 codes,
respectively) and a slight decrease in the mean number performed by PGY-4
residents (147.5 codes). The reason for these slight changes is unclear, but
there appears to be more of a balance in the number of cases performed among
the training years following the adoption of the ACGME regulations. The larger
increase in the number of cases performed by residents at the PGY-5 level
compared with those done by the junior residents may indicate that the slight
changes in our program occurred primarily within the senior resident group.
One rotation change for the PGY-4 residents coincided with the adoption of the
ACGME regulations and may explain the slight decrease in the procedures
recorded by these residents. Just prior to July 1, 2003, we added a new hand
surgeon to our full-time faculty. The PGY-4 hand rotation had previously been
spent working with two established hand surgeons in private practice.
Following July 1, 2003, the PGY-4 resident began splitting time between one of
the surgeons in private practice and the new full-time faculty member. It is
possible that this change led to a decrease in the number of procedures
performed by that resident during this time-period.
In this study, we relied on CPT codes as an indicator of surgical
experience. In our opinion, the number of procedures better represents
resident operative volume than the number of individual cases, many of which
generate multiple procedures. However, an attempt was made to obtain case
numbers for each resident in the study to assess whether there were any
differences between these two measures. Unfortunately, archived data on the
residents did not contain sufficient information, and therefore we had
incomplete data on which to further assess the procedure-case relations hip.
The ACGME utilizes CPT codes as its measure of surgical experience through the
resident case-log database under the report heading of "resident
experience." The CPT codes per resident are organized into ten
categories of procedures, divided into twelve different anatomic sites. The
total number of CPT codes given is the sum of all codes entered into the
database, duplicate and unique.
Determining the number of unique CPT codes performed by each resident may
provide a better understanding of the breadth or quality of the surgical
experience. Unfortunately, acquiring the number of unique CPT codes through
the ACGME database requires looking at the CPT codes in each category and
anatomic site. One hundred and twelve separate groupings are available and are
organized as described above. Each grouping has a large number of unique CPT
codes available. A review of one resident's logged CPT codes during one year
of the study (one resident work year) generated ten pages of individual CPT
codes.
An obvious weakness of this study is the reliance on resident
self-reporting of CPT codes. However, the similar range of codes reported (a
total of 116 to 923 before and a total of 133 to 879 after the start of the
eighty-hour workweek) seems to indicate that there may be a similar number of
low and high reporters per year throughout the years reviewed. At our program,
residents are trained to use the ACGME resident case-log system at the
beginning of their second postgraduate year. They are required to maintain
their case log throughout the year and are encouraged to log cases frequently.
Our residency coordinator checks compliance periodically. Residents who have
not entered cases during their rotations are contacted and are required to
update the case log. We do not currently have a system in place to determine
the accuracy of resident reporting. Determining accuracy would be difficult
with an average annual number of 9100 procedures logged by our residents.
Although it was not significant, the increase in resident operative
experience indicated by the slightly increased number of CPT codes following
the enforcement of the eighty-hour workweek may be secondary to an overall
increase in the number of adult orthopaedic trauma cases at our primary
institution, which became a level-I trauma center in 2001. During the first
two years of the study (July 1, 2001, to June 30, 2003), there were 4964
trauma patients treated at our level-I trauma center. This number increased to
5200 during the second two years of our study (July 1, 2003, to June 30,
2005). We increased our full-time faculty by one surgeon at one year before
the work-hour regulations took effect. It is possible that, with time, the
small increase in our surgical staff led to an increase in the number of
procedures performed by our faculty during the study and, therefore, an
increase in surgical opportunities for the residents. During the study period,
there was no indication that any of the attending surgeons performed any
procedures without resident assistance, as we did not hire any surgical
assistants or physician assistants to help with the procedures.
In maintaining a consistent volume of operative experience after the
implementation of the eighty-hour workweek rules, it is possible that our
residents were spending less time in the outpatient setting. We did not track
the time that our residents spent in clinic during the time-period of our
study; therefore, we have no method for determining whether there had been a
decrease in the clinic experience. Following the introduction of the
eighty-hour workweek, we hired one physician assistant to assist with clinical
care in our pediatric clinics. The effect of this on resident clinical
experience is thought to be minimal as the residents on the pediatric service
do not take in-house call and therefore rarely miss clinics because of
work-hour limitations. At our institution, we generally create schedules that
prevent residents from being post-call on their operative days. Therefore,
they are typically post-call on their clinic days. We do not have a
night-float type of call coverage at our institution, and findings in our
study may not be applicable to programs that employ a night-float system.
Instead, we utilize a combination of fewer in-house-call residents and more
at-home-call residents to comply with the eighty-hour work-week guidelines.
Our system of call coverage involves greater cross-covering of services and
therefore most closely resembles the "stretch model" of call
coverage described by Darosa et
al.11.
Regarding the educational experience in our program, we monitor resident
evaluations of rotations following each rotation approximately five times
annually. There have not been any serious concerns expressed regarding changes
in the educational experience following the introduction of the eighty-hour
workweek. We have not assessed resident satisfaction by any standardized
method. We did not make any specific changes to the resident rotations to
increase the number of operating days in order to provide more operative
experience.
There were no records regarding resident work hours in our program prior to
July 2003. We currently have a computerized system utilizing resident
self-reporting to track work hours for all residency programs at our
institution. It is plausible that our residents were in compliance with the
current work-hour regulations prior to their adoption. We believe it is
unlikely that this was the case as a number of changes have been made in our
resident rotations and call schedules to address issues related to excessive
work hours. These issues became apparent with our current monitoring of our
digital work-hour log. Some of the changes involved adding call
responsibilities to rotations that previously did not have any call
coverage.
An alternative method to determine adequate operative experience would be
to measure the actual time spent on operative training. Chung calculated the
number of hours spent weekly (16.5 hours) on operative training by surgery
residents prior to the introduction of the eighty-hour
work-week12.
Although measuring the actual time spent operating may be a valid measure of
operative experience, it would be much more difficult to track the operative
hours for each resident than to count CPT codes.
In summary, our study indicates that the ACGME work-hour restrictions have
not decreased the experience of orthopaedic residents at our institution. The
use of CPT codes as an objective measure of operative experience is not
without flaws, but it provides a method for individual residency training
programs, and perhaps for the ACGME, to quantify an individual's surgical
training. We believe that we have been able to maintain a consistent operative
experience primarily through adjustments in the type and quantity of call
coverage. A larger, multi-institutional study may provide more strength to
these findings from a single training program. Although we are aware of no
similar published studies on the experience of orthopaedic residents, a number
of studies on general surgery residents have also indicated that there was no
change in operative experience following the induction of the ACGME duty-hour
mandate4,7-9.