The 80-Hour Work Guidelines and Resident Survey Perceptions of Quality

Journal of Surgical Research 135, 275–281 (2006) doi:10.1016/j.jss.2006.04.010

The 80-Hour Work Guidelines and Resident Survey Perceptions of Quality C. Katarina Biller, M.D., M.P.H.,* Anthony C. Antonacci, M.D., S.M.,*,† Stephen Pelletier, Ph.D.,‡ Peter Homel, Ph.D.,§ Cyril Spann, M.D., S.M.,¶ Michael J. Cunningham, M.D.,储 and Roland D. Eavey, M.D., S.M.储,1 *Department of Surgery, Beth Israel Medical Center, New York, New York; †Weill Medical College of Cornell University, New York, New York; ‡Department of Education Development, Harvard Medical School, Boston, Massachusetts; §Department of Pain Management and Palliative Care, Beth Israel Medical Center, New York, New York; ¶Crawford Long Medical Center, Department of Obstetrics and Gynecology, Emory University School of Medicine, Atlanta, Georgia; and 储Pediatric Otolaryngology Service, Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, Department of Otology and Laryngology, Harvard Medical School, Boston, Massachusetts Submitted for publication December 19, 2005

Objective. We hypothesized that resident fatigue error should improve, related to well-rested trainees as a direct cause/effect benefit. However, patient hospital care quality is multifactorial, so impact on patient care quality by changing only one variable for a single caregiver group was unknown. Design and participants. Convenience samples of 156 residents from three surgical specialties were administered a questionnaire in early 2004 addressing perceptions of patient care quality before and after the 80-h workweek. Additionally, residents recently under work-hour restrictions (Newly Restricted, NR) were compared to New York state trainees already regulated by work-hour restrictions (Previously Restricted, PR). Setting. Surgical residency training venues. Main outcome measure. Survey results; the level of significance for all tests was 0.05. Results. The participation response rate was 94.5%. Eighty-eight percent of respondents indicated by survey subjective impression that patient care quality was either unchanged (63%) or worse (26%) due to work-hour restrictions (P ⴝ 0.003). PR residents were more likely than NR residents to report unchanged or worse quality of care (P ⴝ 0.015). Residents overall did perceive improvement in some types of error with fewer fatigue-related errors (P < 0.001), e.g., medication (P < 0.001), judgment (P ⴝ 0.001), and dexterity (P ⴝ 0.013), subsequent to work-hour restrictions. However, more errors were perceived related to con1 To whom correspondence and reprint requests should be addressed at Massachusetts Eye and Ear Infirmary, 243 Charles Street, Boston MA 02114. E-mail: [email protected].

tinuity of care (P < 0.001), miscommunication (P ⴝ 0.001), and cross-coverage availability (P ⴝ 0.001). Conclusions. Despite an expected perception of improvement in fatigue-related errors, most participants (particularly PR residents) reported impressions that patient care quality had remained unchanged or had declined under the work-hour restrictions. Unresolved challenges with continuity of care, miscommunication, and cross-coverage availability are possible explanations. Mere work-hour reduction does not appear to improve patient care quality automatically nor to decrease the possibility for some types of error. Process interventions that specifically target trainee sign-out coverage constraints as part of a global reassessment will be important for future attempts to enhance quality hospital patient care. © 2006 Elsevier Inc. All rights reserved.

Key Words: 80 hour workweek; surgical residents; ACGME work guidelines; quality; medical error; continuity of care; miscommunication; cross coverage; work hour reduction. INTRODUCTION

The debate over resident duty hours reached a pivotal point in 2003 with nationwide institution of the 80-h resident workweek. Although not strictly enforced, precedent for such guidelines previously had existed in New York State, which had limited resident work hours following the death in 1984 of a patient, Ms. Libby Zion, at a New York City teaching hospital. Ms. Zion’s death, attributed to errors stemming from resident fatigue and inadequate supervision, resulted

275

0022-4804/06 $32.00 © 2006 Elsevier Inc. All rights reserved.

276

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

in a set of regulations (the 405 Regulations) that limited resident work hours in July, 1989 [1]. Nationwide extension of such guidelines occurred on July 1, 2003, when the Accreditation Council of Graduate Medical Education (ACGME) mandated that all U.S. residency programs limit work hours for trainees [2]. Under these guidelines, residents are limited to a maximum of 80 work hours per week; in-house call is restricted to every third night; a 10-h work-free period is required after a 24-h shift; and 1 in every 7 days must be free of clinical duties. The limit of 80 h had been calculated arbitrarily; no pilot experiment had been performed to our knowledge [3]. The ACGME released a summary of achievements 12 months after the nationwide implementation of these guidelines [4]. In this summary, achievements related to compliance and organization were emphasized. No evaluation of the perceived or measured impact of resident work hours on patient care quality was mentioned, despite the premise that quality care enhancement was a fundamental rationale for the implementation. Because the 405 Regulations have been in effect for 16 years in New York, earlier investigations of patient care under the 80-h workweek have focused on New York State programs. A 2002 survey of residents in a New York City hospital revealed that such work-hour limitations resulted in a salutary effect on perceived quality of life and basic education for surgical residents, but potentially at the expense of quality and continuity of patient care [5]. A 2003 statewide survey of New York general surgery residents revealed implementation of resident work-hour limitations in general surgery residencies might have negative consequences for patient care and resident education [6]. Surgical residents surveyed and reported in 2004 at four State University of New York programs that, despite decreased fatigue at work, the rate of medical errors had not decreased, and potential adverse effects on patient continuity and safety of care might have resulted from such work-hour limitations [7]. Petersen et al. found that patients’ risk for potentially preventable adverse events were increased when the patients were cross-covered by physicians from another team, particularly when the cross-covering physician was an intern [8]. In a retrospective cohort study of internal medicine teams, Laine et al. found that in-hospital complications and diagnostic test delays were more frequent after enactment of the New York State 405 Regulation, most likely due to crosscoverage by residents unfamiliar with the patients [9]. Given the ambiguous impact work-hour restrictions might produce on patient care, we sought to ascertain the perceived consequences of the regulations from residents in surgical-based specialties who had trained both before and after the national regulations had been introduced. The general surgical residents also pro-

vided two types of comparison since the subpopulation of New York state trainees previously had functioned in a work-restricted environment. The survey made specific inquiries about experiences necessary for optimal patient care; the activation timing of the regulations permitted a unique moment to capture trainee perceptions. METHODS Survey Instrument Development An anonymous survey was designed to measure the subjective impressions of residents about the quality of patient care before and after the 2003 ACGME guidelines. Initial drafts of the questionnaire were developed by two investigators (R.D.E., A.C.A.) at the Harvard School of Public Health. The drafts subsequently were revised by an investigator with expertise in survey design from the Harvard Opinion Research Program (S.P.). The instrument was a 6-page, 63-question inquiry that required approximately 18 minutes to complete. Topics included general impressions about the quality of care provided to patients (14 statements); specific types of clinical training experiences for the resident regarding preoperative, outpatient, and surgical experience (19 statements); types of potential errors such as diagnosis, technique, judgment, communication, supervision, and those related to fatigue (9 statements); routine activities (7 statements); theoretical optimal design of a residency program (4 multiresponse questions); and demographic information (10 questions). As a part of the survey, residents were asked to report when their program adopted the 80-h workweek in relation to July 1, 2003 and the number of hours currently worked per week. The tone of the questions was balanced and varied to avoid unintentional bias due to phrasing connotation. The format was primarily of Likert scale and multiple choice questions.

Survey Population and Survey Administration The survey population was a convenience sample of residents whose care and training had been directly impacted by the 80-h workweek regulations. Persons from three types of surgical training programs were surveyed: (1) general surgery resident attendees of the Third Annual American Board of Surgery In-Training Examination Review Course at Beth Israel Medical Center in New York City, which included residents from several states; (2) Emory University obstetrics/gynecology residents; and (3) Harvard University otolaryngology residents. The surgery and otolaryngology groups were chosen based on the respective programs of the survey investigators, and the OB/GYN group was chosen by program director interest in the survey. The questionnaire was distributed to the general surgery survey population in January, 2004, just over 6 months after the ACGME guidelines took effect. The survey was distributed to the other two groups a few weeks later in early 2004. As an anonymous instrument, residents could elect not to participate.

Statistical Analysis For the analysis of differences between independent groups involving Likert scales (e.g., PR residents versus NR residents with regard to general impression of patient care), the evaluation was carried out using the nonparametric Mann–Whitney test for ordinal differences between independent groups. For comparison of independent groups involving categorical scales (e.g., “yes”, “no”, “don’t know”), the ␹2 test was used. The nonparametric Wilcoxon test for ordinal differences between paired samples was used to compare paired samples involving Likert scales paired data (before the 80-h workweek versus after the 80-h workweek), while the McNemar test

277

BILLER ET AL.: THE 80-HOUR WORK GUIDELINES was used to compare paired samples involving categorical data. The McNemar test was only possible in the case of symmetric cross tabulations (i.e., same number of rows and columns). The level of significance for all tests was 0.05. Data analysis was carried out jointly by Cogent Research and by P.H. [10].

RESULTS Demographic Data

Of the 165 residents to whom questionnaires were distributed, 156 (94.5%) responded, creating a census analysis. Among the 156 respondents, 111 (71.2%) were general surgery residents, 31 (19.9%) were obstetric/ gynecology residents, and 14 (9.0%) were otolaryngology residents. Among the total respondents, 23% were postgraduate year (PGY) 1, 24% were PGY2, 20% were PGY3, 20% were PGY4, and 10% were PGY5; 3% did not respond to the PGY level question. Residency locations included New York (45% of respondents), Georgia (19%), New Jersey (17%), Massachusetts (9%), Pennsylvania (4%), and New Hampshire (1%); 5% did not respond to state of residency. Forty percent of respondents were female, and 83% were age 25–34 years. Work Hours

Seventy-five of 156 (48%) of respondents reported a date prior to the national initiation of ACGME Regulations on July 1, 2003 as the date of official adoption of the 80-h workweek for their program. This group will hereafter be referred to as Previously Restricted residents (PR); all PR residents were from the general surgery subpopulation. Seventy-three of 156 (47%) respondents reported July 1, 2003 as the date of official adoption of the 80-h workweek for their program. As these residents are those new to the ACGME Regulations, they will hereafter be referred to as Newly Restricted (NR) residents. NR residents are summed in the total results; direct comparisons of PR to NR results were between general surgery subpopulations. Five percent of respondents reported no reduction in work hours had occurred in their specific training program. Eighty-two percent of residents report working between 71 and 90 h weekly, a range likely within compliance as the ACGME stipulates a maximum of 80 work hours per week averaged over a 4-week period [11]. Delivery of Quality Patient Care

The majority of residents (89%) responded that the quality of patient care either had gotten worse (26%) or stayed the same (63%). The distribution of responses to this question was significantly different from chance or equal likelihood (P ⬍ 0.001). In addition, PR residents were significantly more likely to report unchanged (61%) or worse (34%) quality of care than NR residents (65 and 17% respectively, P ⫽ 0.003) (Table 1).

A comparable majority of residents maintained that stress levels while on-call were either unchanged (55%) or greater (12%) and the distribution of responses was also significantly different from chance. However, the responses were similar for PR and NR residents and not statistically different (P ⫽ 0.43). Of the 33% who reported less stress, 50% attributed that feeling to more time to rest and sleep. Of the 12% who reported more stress, 50% attributed that feeling to the necessity of covering more services per call shift and 38% to more work. When asked on a Likert scale about whether or not the ability to provide patient care while on-call had been compromised, the respondents split equally with 52% expressing concern and 48% not feeling concerned (P ⫽ 0.62). There were also no significant differences between PR and NR. On another Likert scale item regarding concern about the quality of surgical training being received, 71% overall were “very or somewhat concerned” (P ⫽ 0.001) and again the PR and NR residents largely showed agreement with each other (P ⫽ 0.64). Patient Care Experience

Overall, comparing NR residents after implementation to before implementation, most reported feeling better rested on duty (P ⫽ 0.001), able to make clearer judgments while on duty (P ⬍ 0.001), and able to comfortably cover more patients (P ⫽ 0.04). There were no significant differences between PR and NR residents with regard to these questions (P ⬍ 0.18), which indicates some degree of agreement between these two groups. Other aspects of patient care showed no significant changes from before to after implementation for the overall sample: care given to admitted, operated upon, and signed out patients (P ⫽ 0.21); feeling stretched too thin when covering patients (P ⫽ 0.79); current level of responsibility being appropriate for level of training (P ⫽ 0.38); and supervisions being appropriate to the level of training (P ⫽ 0.18). PR and NR residents for the most part showed no significant differences in their pattern of response to these items, with the exception of two issues. With regard to care TABLE 1 Impression as to Whether the Quality of Patient Care Had Changed After the Start of the 80-h Work Week

Question

Overall

Previously restricted residents

Quality of patient care Better Same Worse

N ⫽ 143 16 (11%) 90 (63%) 37 (26%)

n ⫽ 74 4 (5%) 45 (61%) 25 (34%)

Newly restricted residents

P value

N ⫽ 69 12 (17%) 45 (65%) 12 (17%)

0.003 — — —

278

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

given to admitted, operated upon, or signed-out patients, the NR residents tended to show a greater degree of agreement with this statement after implementation (P ⫽ 0.002). In contrast, for the item dealing with supervision, the NR residents tended to show a slightly lesser degree of agreement with this statement (P ⫽ 0.01). Errors

As seen in Table 2, overall, residents indicated that there was significant improvement after implementation for errors that involved cognitive capacity such as errors involving medication (P ⬍ 0.001), laboratory data evaluation (P ⫽ 0.003), manual technique (P ⫽ 0.013), and judgment (P ⫽ 0.001). Fatigue-related errors also improved (P ⬍ 0.001) as well as errors among colleague staff (P ⫽ 0.001). In contrast, errors involving adequate communication significantly worsened (P ⫽ 0.001) as did errors of

continuity of care (P ⬍ 0.001). Errors that involved inability to be at more than one place at the same time were also increased (P ⫽ 0.001). Errors involving laboratory data evaluation or diagnosis appear to show an ambiguous change since there is a decrease in the “don’t know” response with increases in both “yes” and “no” responses, with a slightly higher increase in the rate of “yes” responses. Comparing PR to NR residents with regard to errors after implementation (Table 3) indicates a higher rate of errors of judgment (P ⫽ 0.02), manual technique (P ⱕ 0.001), errors in communication (P ⫽ 0.01), and continuity of care (P ⫽ 0.045). Routine Activities and Resident Life

After ACGME guideline implementation, residents reported having more time for educational and admin-

TABLE 2 Reported Errors Among All Residents Before and After ACGME Guideline Implementation Type/source of error Laboratory diagnosis (N ⫽ 123) Yes Don’t know No Manual technique (N ⫽ 126) Yes Don’t know No Judgment (N ⫽ 122) Yes Don’t know No Continuity of care (N ⫽ 123) Yes Don’t know No Communication (N ⫽ 121) Yes Don’t know No Coverage (N ⫽ 122) Yes Don’t know No Colleague staff (N ⫽ 125) Yes Don’t know No Medication (N ⫽ 129) Yes Don’t know No Fatigue (N ⫽ 121) Yes Don’t know No a

P value by McNemar test.

Before implementation

After implementation

42 (34%) 44 (36%) 37 (30%)

52 (42%) 29 (24%) 42 (34%)

40 (32%) 37 (29%) 49 (39%)

41 (33%) 24 (19%) 61 (48%)

46 (38%) 35 (29%) 41 (34%)

44 (36%) 24 (20%) 54 (44%)

39 (32%) 32 (26%) 52 (42%)

61 (50%) 18 (15%) 44 (36%)

45 (37%) 33 (27%) 43 (36%)

66 (54%) 18 (15%) 37 (31%)

48 (39%) 36 (30%) 38 (31%)

66 (54%) 19 (16%) 37 (30%)

43 (34%) 35 (28%) 47 (38%)

48 (38%) 18 (14%) 59 (47%)

35 (27%) 52 (40%) 42 (33%)

34 (26%) 27 (29%) 58 (45%)

53 (44%) 35 (29%) 33 (27%)

29 (24%) 25 (21%) 67 (55%)

P value 0.003 a

0.013

0.001

⬍0.001

0.001

0.001

0.001

⬍0.001

⬍0.001

BILLER ET AL.: THE 80-HOUR WORK GUIDELINES

TABLE 3 Comparison of Error Perception Among the PR and NR Groups of Residents After Guideline Implementation

Type/source of error

Overall

Previously restricted residents

Laboratory diagnosis Yes Don’t know No Manual technique Yes Don’t know No Judgment Yes Don’t know No Continuity of care Yes Don’t know No Communication Yes Don’t know No Coverage Yes Don’t know No Colleague staff Yes Don’t know No Medication Yes Don’t know No Fatigue Yes Don’t know No

N ⫽ 85 38 (45%) 17 (20%) 30 (35%) N ⫽ 87 30 (34%) 16 (18%) 41 (47%) N ⫽ 82 31 (38%) 15 (18%) 36 (44%) N ⫽ 85 48 (56%) 10 (12%) 27 (32%) N ⫽ 85 52 (61%) 10 (12%) 23 (27%) N ⫽ 85 48 (56%) 13 (15%) 24 (28%) N ⫽ 86 35 (41%) 13 (15%) 38 (44%) N ⫽ 90 28 (31%) 23 (26%) 39 (43%) N ⫽ 85 20 (24%) 17 (20%) 48 (56%)

n ⫽ 50 27 (54%) 7 (14%) 16 (32%) n ⫽ 50 24 (48%) 12 (24%) 14 (26%) n ⫽ 45 23 (51%) 8 (18%) 14 (31%) n ⫽ 48 35 (73%) 4 (8%) 9 (19%) n ⫽ 48 36 (75%) 4 (8%) 8 (17%) n ⫽ 47 30 (64%) 7 (15%) 10 (21%) n ⫽ 50 23 (46%) 7 (14%) 20 (40%) n ⫽ 52 16 (31%) 13 (25%) 23 (44%) n ⫽ 50 12 (24%) 13 (26%) 25 (50%)

a

Newly restricted residents N ⫽ 35 11 (31%) 10 (29%) 14 (40%) N ⫽ 37 6 (16%) 4 (11%) 27 (73%) N ⫽ 37 8 (22%) 7 (19%) 22 (59%) N ⫽ 37 13 (35%) 6 (16%) 18 (49%) N ⫽ 37 16 (43%) 6 (16%) 15 (41%) N ⫽ 38 18 (47%) 6 (16%) 14 (37%) N ⫽ 36 12 (33%) 6 (17%) 18 (50%) N ⫽ 38 12 (32%) 10 (26%) 16 (42%) N ⫽ 35 8 (23%) 4 (11%) 23 (66%)

P value 0.09 a

⬍0.001

0.02

0.002

0.01

0.24

0.50

0.98

0.21

P value by ␹ 2 test.

istrative duties. Over 88% of residents felt that they are on the trajectory to achieve the six basic ACGME competency requirements. Theoretical Design of Residency

When asked to list five changes that could be instituted for a residency program to optimize patient care, the most popular responses were as follows: competent ancillary staff (56%); committed attending physicians and fellows (47%); better teaching rounds (47%); more surgical experience (45%); and electronic medical lab reports (37%). Changes that only few residents deemed important included: fewer patients to cover (6%); an additional year of training (3%); more time to interact

279

with patients (3%); more time to round on patients (3%); and additional medical rotation experience (3%). DISCUSSION

The resident workweek was limited to 80 h with a laudable goal of enhanced patient care quality. However, the nationally implemented schedule change, presumed beneficial due to anticipated fatigue relief for vital trainee caregivers, had not been studied prospectively as to actual impact [3]. Conscientious concerns had even been raised about potential negative unintended consequences for patients, especially regarding on-call coverage and professionalism [11]. Intuitively, limiting resident work hours should benefit current patient care by reducing errors related to resident fatigue as well as creating time for learning to benefit future patients [12]. The survey results demonstrate that residents do perceive that several types of error indeed have improved regarding medication use (e.g., wrong dose, wrong route, wrong patient), manual technique (e.g., ability to start an IV, avoidance of pneumothorax during central line placement, surgical mishap), judgment (e.g., underappreciation of patient illness, incorrect surgical decision), and fatigue (e.g., slept through a page, miscalculation of medication dosage). However, although more residents now report feeling better rested while on duty since the implementation of work-hour regulations, paradoxically a substantial proportion reported the perception of unimproved (63%) and even negative (25%) impacts on patient care quality. Only 12% of respondents felt that patient care quality actually had improved, which suggests that the main goal of patient safety has not been achieved. Further, negative perceptions of the impact of the duty hour restrictions were particularly prevalent among PR residents (those under New York State 405 Regulations before the nationwide 2003 institution of the 80-h workweek). This finding suggests that even after functioning for many years in a work–time-restricted environment that quality patient care does not evolve automatically. Reinforcing the finding of negative impact on patient care in this study, a similar negative perception was noted by 43% of surgical trainee respondents in another recent survey [13]. The current survey identifies specific deficiencies in workflow processes that systematically increase the opportunity for error—i.e., miscommunication, poor continuity of care, and physical limitation to cross-coverage issues—seemed to negate presumed benefit from work-hour reduction. Improving quality of care requires systematic review of the causes of error and development of methods for carefully identifying factors that lead to error. To comply with the July 2003 implementation, many programs used schedule changes and night float systems to bring duty hours below the

280

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

acceptable limit. Thus, on-call teams might cross-cover an increasing number of unfamiliar patients. Like other studies [8, 9], our results indicate that errors significantly correlate with cross-coverage by physicians who are less familiar with the patients than the usual care providers. In addition, the data suggest that in both the PR and the NR groups errors based on inadequate continuity of care and miscommunication increased after the implementation of the national work reduction. Further, the perception of being unable to cover patients as well after work-hours reduction is reinforced with the finding that NR residents report that the care provided to sign-out patients is not equal to the care provided to patients personally operated upon or admitted (P ⫽ 0.002). The results are reinforced by the perceptions of another survey of surgical trainees that reported that 32% felt that the risk of patient management errors had increased and that “70% perceived decrements in the continuity of care” [13]. The current level of hospital supplemental system response to hire support staff or to improve computer and communication systems apparently is not filling the gaps required by the work restrictions. Nearly half of residents in both PR and NR groups perceived that hospitals have not hired additional nurses, physician’s assistants, or other support staff to provide services once performed by the residents. Thus, although residents’ working hours are decreased, the expectation exists to perform the same on-call tasks with fewer residents. In 1989 when the New York State 405 Regulations were put into effect, hospitals received $1.2 billion over 7 years to assist in the work-hour compliance [1]. What process implementations might be targeted for future improvement? The overall results demonstrate that continuity of care, miscommunication, and inability to provide coverage at more than one site simultaneously remain unsolved— even exacerbated— problems. The common theme involves coverage duties. An alternative technique to enhance information transfer to reduce errors could be use of technologies such as hand-held devices or laptop computers. A prospective, randomized, crossover study showed that a well-designed and institutionalized computer system for rounding and sign-out enhanced patient care by improving resident-reported quality of sign-out communication and continuity of care, among other positive outcome measures [14]. Additionally, drug-related adverse events can be reduced by the use of algorithms incorporated with computerized physician order entry technology. In this study, new sign-out communication systems had been initiated at the minority of programs, either formal (27%) or informal (49%). However, we did not attempt to measure a correlation as to whether the systems achieved improved effective-

ness. Further, employment of hospital personnel such as hospitalists or physicians’ assistants must be considered, balancing immediate financial implications, because poorer care quality will result eventually in increased costs. Importantly, a new mindset incorporating a “culture of quality” must be consciously pursued and we physicians must take the responsibility to improve patient care as quality involves more than house officer training and error monitoring [15]. Effective implementation of surgical house officer workhour reduction has required attention to practical, cultural, and political according to a recent ethnographic analysis [16]. Additional proactive and systemic suggestions have been detailed previously to assist our colleagues-in-training [17]. The workweek restrictions are too recent to provide many examples of published systems adjustment; however, one surgical trauma service deployed a policy of direct admissions to orthopedic and neurosurgical services. The results demonstrated no adverse measured patient outcome [18]. The study used a survey instrument focused primarily on quality of patient care. The timing of the survey was selected to capture current impressions as well as recent memory about patient care from two time-contrasted environments so that the residents could serve as internal comparisons. Additionally, the unique nature of the New York state general surgery residents allowed a separate comparison of PR versus NR systems and the trends remained similar despite the less robust statistical power of the latter comparison technique. The extremely high response rate (94.5%) allowed even for interpretation as census data. Surgical residents are a reasonable to group to evaluate due to the challenges of care delivery via multiple venues, medical care on patient floors, as well as surgical care in the operating room. A major limitation of this survey is that the data were not obtained as a random sample and instead from three convenience groups. Fortunately, training programs are required to have significant uniformity by the ACGME and no obvious reason exists to think that these survey residents are in training programs that might not represent national standards that could be extrapolated to a larger trainee population. A further limitation is that the survey focused on only one subpopulation in the health care quality picture, surgical house officers, and did not query other impacted groups such as medical residents, nursing staff, and patients. Additionally, since a survey technique was used, objective outcome data were not harvested to determine if respondent impressions correlated with actual patient care quality results or with hospital systems responses. CONCLUSIONS

Despite a significant subjective perception of decline in fatigue-related errors, most surgical residents sur-

BILLER ET AL.: THE 80-HOUR WORK GUIDELINES

veyed felt that quality of patient care had not improved or had even declined under the ACGME work-hour restrictions. Difficulties in continuity of care, miscommunication, and cross-coverage physical limitations are suggested workflow process explanations that seem to negate the benefit of work-hour reduction and to exaggerate coverage types of error. Yet those same findings illustrate specific situations that can be addressed as one component of hospital management to improve future care. The results further suggest that improved patient care quality is not related to long-term experience with a reduced workweek system, as the New York State general surgery residents previously under similar guidelines had significantly and paradoxically worse perceptions about care quality. Continued refinement of quality systems and resident education is essential, and longitudinal study of care quality will be needed to monitor actual objective errors and outcomes.

5. 6.

7.

8.

9.

10. 11.

ACKNOWLEDGMENTS

12.

No author has a financial interest in the results of this publication. Funding was obtained from the Massachusetts Eye and Ear Infirmary, Boston, and the Beth Israel Medical Center, New York, for survey design and for data acquisition and analysis.

13.

14.

REFERENCES 1.

Wallack MK, Chao L. Resident work hours: the evolution of a revolution. Arch Surg 2001;136:1426. 2. Accreditation Council for Graduate Medical Education. Report of the ACGME Work Group on Resident Duty Hours [Accreditation Council for GME web site]. Available at http:// www.acgme.org/new/wkgreport602.pdf. Accessed June 11, 2002. 3. Bell B. Reconsideration of the New York State laws rationalizing the supervision and the working conditions of residents. Einstein J Biol Med 2003;20:36. 4. Accreditation Council for Graduate Medical Education (ACGME).

15.

281

The ACGME’s Approach to Limit Resident Duty Hours 12 Months After Implementation: A Summary of Achievements. September 2004. Barden CB, Specht MC, McCarter MD, et al. Effects of limited work hours on surgical training. J Am Coll Surg 2002;195:531. Whang EE, Mello MM, Ashley SW, et al. Implementing resident work hour limitations: lessons from the New York State experience. Ann Surg 2003;237:449. Kort KC, Pavone LA, Jensen E, et al. Resident perceptions of the impact of work-hour restrictions on health care delivery and surgical education: time for transformational change. Surgery October 2004;136:861. Petersen LA, Brennan TA, O’Neil AC, et al. Does housestaff discontinuity of care increase the risk of preventable adverse events? Ann Intern Med 1994;121:866. Laine C, Goldman L, Soukup JR, et al. The impact of a regulation restricting medical house staff working hours on the quality of patient care. JAMA 1993;269:374. Cogent Research, 2 Tyler Court, Cambridge, MA 02140. www. cogentresearch.com. Accessed May 30, 2006. Lockwood CJ. The 80 hour resident work week: cure or curse? Contemp Ob/Gyn 2003;2:8. Gaba DM, Howard SK. Fatigue among clinicians and the safety of patients. N Engl J Med 2002;347:1249. Irani J, Mello M, Ashley S, et al. Surgical residents’ perceptions of the effects of the ACGME duty hour requirements 1 year after implementation. Surgery 2005;138:246. Van Eaton EG, Horvath KD, Lober WB, et al. A randomized, controlled trial evaluating the impact of a computerized rounding and sign-out system on continuity of care and resident work hours. J Am Coll Surg 2005;200:538. Brennan TA. Physicians’ professional responsibility to improve quality of care. Acad Med 2002;77:973.

16.

Kellogg KC, Breen E, Ferzoco SJ, et al. Resistance to change in surgical residency: an ethnographic study of work hours reform. J Am Coll Surg 2006;202:630.

17.

Volpp KGM, Grande D. Residents’ suggestions for reducing errors in teaching hospitals. N Engl J Med 2003;348:851.

18.

Rogers F, Shackford S, Daniel S, et al. Workload redistribution: a new approach for the 80 hour work week. J Trauma 2005;58: 911.