THE ACCREDITATION COUNCIL for Graduate MedicalEducation (ACGME) and the Surgery ResidencyReview Committee have recently mandated strictregulation of resident work hours.1 These regula-tions were enacted to ensure that patients weretaken care of by physicians who were not sleep-deprived, and that residents were well rested andhad ample time for personal lives. The Society ofUniversity Surgeons (SUS), in its statement on sur-gical resident work hours and education, stated thatthe emphasis on hours worked [has] diverted atten-tion from working environment, at the same timethat working conditions. . . for surgical residentshave been changing, resulting in substantiallyincreased. . . workload and stress for residents.2
Improving the resident work environment shouldnot be limited solely to limiting the amount of timespent in the hospital. It is clear that residents spend a
significant amount of time doing work that is neithereducational nor directly related to patient care. Inaddition, even patient care activities can be redun-dant or inefficient, adding unnecessarily to the resi-dents workload. Rather, all aspects of the residentwork environment should be evaluated to find addi-tional ways to increase sleep and study time, toimprove efficiency, and decrease time-wasting activi-ties while still maintaining excellent patient care. Theimpetus behind this study was to find ways to improvethe resident work environment during duty hours.Time spent answering pages can be significant3,4 andcan interfere with time spent on more urgent patientcare or educational activities or even with sleep. Thepurpose of this study was to try to develop a programthat would improve communication between nursesand residents while decreasing the number of nonur-gent and unnecessary pages that the interns received.A secondary goal of the program was to consistentlygenerate a 4-hour block of time at night that was freeof nonurgent pages that could be used by on-call res-idents for sleep or study time.
MATERIAL AND METHOD Stanford Hospital is a 450-bed university hospi-
tal. Non-intensive care unit surgical beds are
Presented at the 64th Annual Meeting of the Society ofUniversity Surgeons, Houston, Texas, February 12-15, 2003.
Reprint requests: Myriam J. Curet, MD, H3680 Department ofSurgery, Stanford Hospital, 300 Pasteur Dr, Stanford, CA 94305.
2003 Mosby, Inc. All rights reserved.
0039-6060/2003/$30.00 + 0
Improving resident work environment:Evaluation of a novel cooperative programMyriam J. Curet, MD, and Timothy R. McAdams, MD, Stanford, Calif
Background. Improving the resident work environment is a major concern for surgery faculty. Thisstudy evaluated the ability of a cooperative program with nurses and interns to decrease the number ofnonurgent pages and consistently generate a 4-hour block of time at night without nonurgent pages.Methods. Multiple discussions with interns and with nurses on 2 nursing floors identified ways toimprove nurse/resident communication. These included use of a notebook by nurses to record nonurgentissues and having on-call interns check with the night nurses after night shift report. For the week beforeand after institution of the program, interns logged each page received. Pretest and posttest data werecompared by use of t testing.Result. Interns logged fewer pages after intervention compared with preintervention ( P < .01). In addi-tion, the interns had a 4-hour block of time on call nights without pages more frequently during theposttest period (100% vs 25%, P < .01). The percent of necessary calls increased from 50% to 70%during day shifts ( P < .01).Conclusions. A cooperative program that focused on decreasing nonurgent pages and maximizing effi-cient communication led to a decrease in the number of nonurgent pages received by interns andincreased the number of call nights in which a 4-hour block of sleep or study time was generated, therebyimproving residents work environment. (Surgery 2003;134:158-63.)
From the Department of Surgery, Stanford Hospital, Stanford, Calif
Surgery Curet and McAdams 159Volume 134, Number 2
spread out over several wards, but most surgicalpatients who are not in the postoperative cardiac,vascular, or thoracic unit, are primarily admitted to2 nursing units, C3 and E3. Resident call forpatients on C3 and E3 is generally every thirdnight. Institutional Research Review Boardapproval of this study was obtained. Informed con-sent was obtained from all interns who loggedpages. Interns were financially rewarded for turn-ing in their page logs.
Over the course of several weeks before this study,the principal investigators met numerous times sepa-rately with the postgraduate year 1 (PGY 1) residentsat Stanford Hospital and with the nurses on the 2study floors (C3 and E3). The meetings focused onfinding ways to improve nurse/resident communica-tion, to decrease pages of a nonurgent nature to thePGY 1 residents, and to minimize nonurgent pagesbetween 1 AM and 5 AM. A cooperative program withspecific instructions for the PGY 1 residents and thenurses was developed on the basis of the concernsexpressed during these meetings (Table I). Thecooperative programs foundation was the use of anotebook by the nurses in which they would recordconcerns of a nonurgent nature. The PGY 1 resi-dents checked the notebook whenever they were onthe floor and when they answered a page. In addi-tion, PGY 1 residents were expected to check withthe night shift nurses after report around 11:45 PM.Other duties, listed in Table I, included better notifi-cation of when the off-call PGY 1 resident signed outto the on-call PGY 1 resident, better notification ofwho the on-call PGY 1 resident was, and better com-munication of newly entered electronic orders.Some duties, such as anticipating discharge planningor finishing work generated by evening work beforesign-out, did not directly involve communicationwith nurses but were instituted to address the nursesconcerns.
The nurses responsibilities centered on determin-ing which of their questions and concerns wereurgent or nonurgent, so they could decide whether apage was warranted or whether the CommunicationNotebook could be used. The meetings between theprincipal investigators and the nurses on the studyfloors reviewed what situations should be consideredurgent versus nonurgent. General guidelines statedthat a change in the patients status warranted a page.Instructions for what laboratory values and radiogra-phy results necessitated a page were also reviewed.Situations regarding patient and family requests wereleft up to the nurses discretion. Nurses were instruct-ed to err on the side of paging rather than using theCommunication Notebook in situations where theywere uncertain if the matter should be classified as
urgent or nonurgent. The nurses were also asked tocommunicate with each other before paging the PGY1 resident so one page could answer several concerns,instead of each concern generating a separate page.Finally, to avoid long waits when answering a page toa back desk phone, all pages were to go through thefront desk where the clerk was instructed on whichnurse initiated the page and which PGY 1 resident waspaged. This also allowed the clerk to review theCommunication Notebook to ensure that issuesentered there were addressed when the PGY 1 resi-dent returned the page.
Data collection was limited to 8 PGY 1 residents(categorical and preliminary) on the StanfordHospital surgical services that admitted primarilyto C3 and E3 (general surgery, transplant, ortho-pedics, otolaryngology, plastic surgery, and trau-ma). For 7 consecutive days before instituting thenew program, the PGY 1 residents logged all pagesthey received. For each page, they noted the dateand time of the page, from what floor the call orig-inated, the reason for the page, and whether theybelieved the call was necessary. The PGY 1 residentswere also asked to record their perception of howoften they slept from 1 AM to 5 AM during call nights(0% to 10% of call nights; 11% to 25% of callnights; 26% to 50% of call nights; 51% to 75% ofcall nights; 76% to 100% of call nights). Nurses onthe study floors and PGY 1 residents were asked torate their satisfaction with nurse/resident commu-nication on a 5-point Likert scale, with 1 being thelowest and 5 being the highest satisfaction.
After review of the new program with both PGY1 residents and nurses on the study floors, the pro-gram was implemented. During the second weekafter implementation, the PGY 1 residents againrecorded the same information for all pagesreceived for 7 consecutive days. PGY 1 residentsagain noted their perception of how often theyslept from 1 AM to 5 AM during call nights. In addi-tion, nurses and PGY 1 residents again rated theirsatisfaction with nurse/resident communicationon the same Likert scale. At the end of the secondstudy week, qualitative comments on the new pro-gram were elicited from nurses on the study floorsand from the PGY 1 residents.
The total number of pages were totaled for day-time hours (6 AM-7 PM), nighttime hours (7 PM-6AM), and from 1 AM to 5 AM. These groups were tab-ulated for the study floors (C3 and E3) and for allother floors. Data were compared between prein-tervention and postintervention groups. The aver-age rating for nurses and PGY 1 residents satisfac-tion with nurse/resident communication wasdetermined by adding together all the responses in
RESULTSEight PGY 1 residents were asked to record data in
the preintervention group and 8 in the postinterven-tion group. Response rates were 100% for bothgroups. The number of pages from the nonstudyfloors did not change significantly from the preinter-vention to the postintervention groups during day-time or nighttime (data not included). The numberof pages from the study floors decreased significant-ly during daytime shift, during nighttime overall, andbetween 1 AM and 5 AM after initiation of the new pro-gram (Table II). The percent of calls that werebelieved to be necessary increased significantly from56% to 69% overall (P < .05). This change was mostpronounced during daytime shift (50% to 70%; P < .01). PGY 1 residents perception of how muchthey slept during call nights increased significantlypreintervention to postintervention (P < 0.05) (Fig 1). Nursing ratings of nurse/resident communi-cation increased from 3.0 to 3.6 (P < .05). PGY 1 res-idents ratings increased from 3.1 to 3.9 (P < .05).
A review by the principal investigators of the rea-sons for the pages found a surprisingly high num-ber of pages that no one answered when the pagewas returned, when the wrong doctor had beenpaged, or when no one knew why the intern hadbeen paged. There were also a number of infor-
160 Curet and McAdams SurgeryAugust 2003
each group, then dividing by the total number ofresponses for each group. Preintervention andpostintervention ratings were compared.
Statistical analysis was performed with t testingor Pearsons 2 test as appropriate comparingpreintervention results with postinterventionresults. A P value
mational pages informing the intern when anadmission had arrived, a preoperative work-up wasneeded, or rounds were starting.
Qualitative feedback from nurses and PGY 1 res-idents was uniformly positive. They all believed thatthe program achieved its aim of improving com-munication between PGY 1 residents and nurses.Nurses and PGY 1 residents alike believed thatpatient care was not adversely impacted with initia-tion of the new cooperative program. PGY 1 resi-dents believed that they got more sleep on call afterinstitution of the new program. PGY 1 residentsbelieved that they noticed a decrease in the num-ber of nonurgent pages they received and adecrease in unnecessary pages. Interns and nursesboth praised the Communication Notebook andthe resident rounding after nightshift report as thekey foundations to the programs success. No neg-ative comments were elicited.
DISCUSSIONMultiple organizations, including the American
College of Surgeons (ACS), the ACGME, and theSociety of University Surgeons, have addressed theissue of resident work hours, stating that residenteducation and patient care should be the first prior-ities.2,5 These organizations realize that efforts toimprove the resident work environment should notfocus only on work hours but should also address thework environment as a whole. Several authors havefocused on eliminating noneducational activity, vari-ously defined as scut work,6 ancillary tasks,7 orpatient care not requiring a physician,3 as a way ofdecreasing resident work hours. In these studies,these activities accounted for a range of 3.4% to 21%of the residents work week.3,6,7 Interestingly, Warneret al3 found that residents actually spend very littletime, less than 1 hour/day, on patient care dutiesthat do not require a physician.
Defining what is patient care versus educationalversus service work is difficult and controversial,with significant overlap of given activities betweendifferent classifications. Pages, for example, can berelated to patient care (Your patient is in atrial fib-rillation), educational (discussing that patientwith the attending), service work (finding out thatthe patient has been in atrial fibrillation for 10years and has had no acute hemodynamic change),or unnecessary (Sorry, this is not your patient).Studies analyzing how residents spend time do notdirectly address the issue of pages.3,6-9 Someauthors classify pages as direct patient care,7 butmost either do not mention them6,8,9 or state thattime spent on answering pages could not be accu-rately recorded.3
The cumulative time spent answering pages maybe great.3 Katz and Schroeder4 found that medicalinterns were paged an average of once per hour,but that only 34% of these pages required anurgent response or resulted in a significant changein patient care. In this study we found that internscould receive as many as 79 pages in 1 day. If halfof these pages are nonurgent and can be eliminat-ed, that intern now has gained a minimum of anadditional half hour that can be used to read on arecent admission, spend more time with thatadmission, or address a personal matter. Or, ifthose pages can be grouped together, the internwould experience fewer interruptions. Frequentinterruptions can hinder patient care, affect thephysician/patient relationship, contribute to stressin the interns life, make the resident feel overbur-dened with the clinical responsibilities, and makethe intern feel like they have little control over theconditions of their work.4 Katz and Schroeder4
believed that reducing the number of unnecessarypages and postponing nonurgent ones could resultin a 42% decrease in disruptions of patient careand more rest for interns.
Making patient care more efficient does requireeffort on the part of ancillary support staff such asnurses. The ACS states that it is inappropriate forteaching hospitals to rely upon residents to per-form tasks that are not directly related to eithereducation or patient care.5 Furthermore, the ACSstates that it is essential that hospitals provide suf-ficient support personnel to perform noneduca-tional tasks.5 The ACGME states that makinggood use of the time and capabilities of all healthcare professions. . . should be a central goal ofongoing reform.1 Better use of nurses, with more
Surgery Curet and McAdams 161Volume 134, Number 2
Fig 1. Responses to: When you are on call on your pre-sent service, how often do you get sleep from 1 AM to 5AM? (y axis = # of responses).
responsibility and independence on their part,would avoid the need to hire additional personnelat a time of critical money shortages. A recent dis-cussion group led by Dr Timothy Flynn, whenasked How can you restructure the patient careteam to better utilize residents and achieve compli-ance? listed increasing the responsibilities of nurs-es and improving nurse/resident/faculty commu-nication (unpublished data). Colleagues in privatepractice who are called directly by nurses withoutresidents as intermediaries, know that nurses arecapable of following clinical pathways, determiningurgency of patient care situations, and makingdecisions about what requires an urgent page andwhat issues can be postponed. For unknown rea-sons, nurses at teaching hospitals typically do notachieve this level of responsibility. Rather, theintern becomes the repository for all questions andissues, regardless of urgency and without concernabout the interns work schedule.
An important part of our cooperative programwas to set guidelines for the nurses on what shouldbe considered urgent versus nonurgent. Katz andSchroeder4 agree with us that communicationbetween residents and nurses is critical to addressappropriate indications for paging and to improvecommunication. Katz and Schroeder4 also believedthat having an alternative to paging for nonurgentsituations such as message boards or a scut sheetwould be helpful, as would grouping nonurgentpages or having nursing supervisors review ques-tions before paging.
In the discussion about resident work hour limi-tations, much attention has been focused on theeffects of sleep deprivation. Effects of sleep depriva-tion in residents are hard to define and measure,and vary from individual to individual. A number ofreports have indicated that sleep deprivation occurswhen individuals sleep less than 3 hours in a 24-hour period.3,10,11 With this definition, generating a4-hour block of time without nonurgent pages thatcould be used for sleep would significantly improveresident function and mood. Other studies havefound that 2 hours of sleep before 24 hours of sleeploss can significantly minimize the effects of sleepdeprivation.12 Even short naps of 15 minutes dura-tion can significantly decrease sleepiness.12 Giventhat the 2 AM to 9 AM period of time is the circadiannadir, and therefore the time period least likely torespond to measures designed to counter sleepi-ness,11,12 generating a window of time for sleepfrom 1 AM to 5 AM would be extremely helpful to res-idents. According to the study by Warner et al,3
nighttime is the time when the least amount ofpatient care activities occur, so efforts to be particu-
larly efficient with pages, to anticipate patientsneeds, and to address them earlier in the nightcould be very successful in making call nights morecomfortable.3
Several unexpected findings came out of ourstudy. Pages were often used to communicate infor-mation to interns that did not require the intern toreturn the page. Pages sent to give information onlaboratory results, to inform the intern that roundsare starting, or that an admission has arrived can allbe transmitted via text pages, thus making moreefficient use of intern time.
An important issue that is not addressed by thedata presented is the sustainability of these results.This study was performed in the spring of 2002,and the PGY 1 residents continued to use theCommunication Notebook and to round afternight shift. A new group of PGY 1 residents wereoriented in July 2002. Anecdotally, we found thatthese new PGY 1 residents needed frequentreminders to check the Communication Notebookwhen they returned to Stanford after rotating atoutside institutions. We also found the programharder to institute on nursing units where bothsurgery and medicine patients were admitted. Theprogram appears to be more sustainable on nurs-ing units with a major focus on surgical patients.The rounding at night appears to be easily sustain-able. We plan to continue to implement the pro-gram on the study floors and to expand as possibleto all floors in the hospital.
CONCLUSIONSThere are numerous ways to improve the resi-
dent work environment. Better use of resident timewith less time spent on nonurgent patient care mat-ters would free up time for sleep, study, and per-sonal time. A successful program to achieve theseaims can be implemented. It is important to listento the concerns of interns and nurses before for-mulating a cooperative program so that everyonesconcerns are met. The use of a CommunicationNotebook and rounding after nightshift reportwere instrumental in establishing successful com-munication between interns and nurses that wasmore efficient without impacting patient care.
REFERENCES1. Philibert I, Friedmann P, Williams WT. New requirements
for resident duty hours. JAMA 2002;288:1112-4.2. Cole DJ, Bertagnolli MM, Nussbaum M. Society of
University Surgeons statement on surgical resident workhours and education. Surgery 2002;132:115-8.
3. Warner BW, Hamilton FN, Brunck BS, Bower RH, Bell RHJr. Study of surgical resident working hours and time uti-lization. J Surg Res 1990;48:606-10.
162 Curet and McAdams SurgeryAugust 2003
Surgery Curet and McAdams 163Volume 134, Number 2
4. Katz MH, Schroeder SA. The sounds of the hospital: pagingpatterns in three teaching hospitals. NEJM 1988;319:1585-9.
5. Statement on residency work hours. Bulletin Am Coll Surg;2002. p. 21.
6. Schwartz RJ, Dubrow TJ, Ross RF, Williams RA, Butler JA,Wilson SE. Guidelines for surgical residents working hours:intent vs reality. Arch Surg 1992;127:778-83.
7. Steer KS, Peoples JB. A study of the on-duty hours of surgi-cal residents. Surgery 1990;108:393-8.
8. Miller SF. Composite resident workweek. Am J Surg1992;164:377-81.
9. Magnusson AR, Hedges JR, Harper RJ, Greaves P. First-post-graduate-year resident clinical time use on three specialtyrotation. Acad Emerg Med 1999;6:939-46.
10. Reznick RK, Folse JR. Effect of sleep deprivation on the per-formance of surgical residents. Am J Surg 1987;154:520-5.
11. Deaconson TF, OHair DA, Levy MF, Lee MBF,Schueneman Al, Condon RE. Sleep deprivation and resi-dent performance. JAMA 1988;260:1721-7.
12. Veasey S, Rosen R, Barzansky B, Rosen I, Owens J. Sleep lossand fatigue in residency training: a reappraisal. JAMA2002;288:1116-24.