A C TA Obstetricia et Gynecologica
AOGS M A I N R E SE A RC H A R TI C LE
Impact of night shifts on laparoscopic skills and cognitive function among gynecologists ANNE VEDDENG1, THOMAS HUSBY2, INGEBORG B. ENGELSEN1, ANDREW KENT3,4 & HANS FLAATTEN2,5 1
Department of Gynecology and Obstetrics, 2Department of Anesthesiology, Haukeland University Hospital, Bergen, Norway, 3Advanced Gynecological Endoscopy Program, University of Surrey, 4Royal Surrey County Hospital, Guildford, UK and 5Department of Clinical Medicine, University of Bergen, Bergen, Norway
Key words Assessment of surgical skills, virtual reality simulator, sleep deprivation, work-hour restrictions, patient safety Correspondence Anne Veddeng, Department of Gynecology and Obstetrics, Haukeland University Hospital, 5021 Bergen, Norway. E-mail: [email protected] Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. Please cite this article as: Veddeng A, Husby T, Engelsen IB, Kent A, Flaatten H. Impact of night shifts on laparoscopic skills and cognitive function among gynecologists. Acta Obstet Gynecol Scand 2014; 93: 1255–1261. Received: 2 November 2013 Accepted: 25 August 2014 DOI: 10.1111/aogs.12496
Abstract Objective. To assess whether gynecologists have impaired laparoscopic skills and/or reduced cognitive function after long on-call hours. Design. Prospective cohort study. Setting. Department of Gynecology and Obstetrics, Norway. Population. 28 gynecologists working long shifts in a single department. Methods. Pre-training of laparoscopic skills on a virtual reality simulator. Simulator- and cognitive testing on two different occasions; one in the morning after a normal nights’ sleep at home and one in the morning directly after 17.5 h on call. The virtual reality simulator test consisted of three repetitive salpingectomies in an ectopic pregnancy module. The cognitive test consisted of a standardized cognitive test battery (Cambridge neuropsychological test automated battery). Main outcome measures. Simulated laparoscopic performance was assessed by time to perform the procedure, total length of instrument movement (tip trajectory) and blood loss. Cognitive function was assessed by reaction time, errors and total score. Results. No significant impairment in laparoscopic skills was found after 17.5 h on call. Cognitive testing revealed a statistically significant increase in reaction time post-call. Construct validity for the metrics “time to perform procedure” and “tip trajectory” in the ectopic pregnancy module was established in a validation study prior to our main study. Conclusions. We were not able to detect impaired laparoscopic skills among gynecologists tested by a virtual reality procedural module after 17.5 h on call. We found a small increase in reaction time but no other signs of reduced cognitive function. The study adds information on surgical performance of sleep-deprived gynecologists. CANTAB, Cambridge neuropsychological test automated battery; EP, ectopic pregnancy; tip trajectory, total length of instrument movement; VR, virtual reality.
Abbreviations:
Introduction The duration of on-call shifts for medical professionals is a matter of debate and has frequently been the focus of discussion in the Norwegian media as well as of the national labor inspection authority. The media suggest that physicians demand long work hours only to increase
Key Message Simulated laparoscopic skills are not necessarily impaired among professionals in a clinical setting at the end of long periods on call. Cognitive testing may add strength to the assessment of surgical skills among sleep-deprived surgeons.
ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 1255–1261
1255
Impact of night shifts on gynecologists
salary, jeopardizing patient safety (1). Norwegian working hour regulations limit physicians to work a maximum of 19 consecutive hours and a maximum of 60 h a week. The European Working Time Directive for physicians in the European Union is even stricter, with a maximum of 13 consecutive hours and 48 h a week (since August 2009). Due to increased focus on patient safety there is a demand for consultants to be available on-site 24/7 in Norwegian hospitals, not at least in the field of obstetrics and gynecology. Physicians are exempted from their daily work before and after being on call, and shorter duration and more frequent on-calls will therefore lead to a lack of consultants to cover specialized tasks in the daytime. Moreover, physicians’ clinical skills are at risk of being impaired, as these are mainly developed and maintained by daytime work. Research has demonstrated that one night of prolonged wakefulness causes performance impairments equivalent to an alcohol intoxication of 0.10% (2). Although there is little doubt that sleep deprivation impairs human functioning, typical surgical skills do not necessarily deteriorate with a limited amount of sleep loss under average conditions in a hospital or clinic (3,4). For obvious reasons, no randomized controlled trials have examined the direct effects of surgeon fatigue on patient outcome, and therefore simulation studies have frequently been used in this field of research (4–9). However, few trials have included surgeons in real clinical settings. Cognitive testing has often been omitted, although it is widely known that surgery consists of good judgment as well as technical skills. Gynecologists/obstetricians on call are prone to sleep deprivation and they often have to perform emergency surgical procedures during their night shifts. Primarily, we aimed to investigate simulated laparoscopic skills and cognitive function among gynecologists immediately after 17.5 h on call as compared to a rested situation. Secondarily, we wished to investigate the construct validity of relevant metrics in the ectopic pregnancy (EP) module in the SimSurgery virtual reality (VR) simulator.
Material and methods Twenty-eight of 33 gynecologists working on-call shifts at the Department of Gynecology and Obstetrics, Haukeland University Hospital, were enrolled in the study between October and November 2011. The reasons for not participating were planned study or maternity leave in the period of trial. All participants gave informed consent under a protocol approved by the Regional Ethic Committee in Norway (REK Vest, 2011/1900). The trial was completed in April 2012.
1256
A. Veddeng et al.
Haukeland University Hospital is a secondary and tertiary health care facility, providing health services to approximately 1 million inhabitants living in the western part of Norway. The annual number of deliveries at the Department of Gynecology and Obstetrics is approximately 5000. The gynecologists working on-call have on average one on-call shift a week (15.30–09.00 h), and they are exempted from regular working hours before and after being on call. Due to the intensive workload, and for educational and safety reasons, gynecologists from three different tiers, a junior trainee, a more experienced trainee and an experienced specialist, stay at the hospital during the whole on-call shift. A minority of the participants had experience with the VR simulator SimSurgery Educational Platform (SimSurgery AS, Oslo, Norway) before the start of this study, but none during the past year. To familiarize the participants with the test tool and to diminish learning bias during the study period, they received pre-study supervised training on the VR simulator, using the SimSurgery Educational Platform Basic and EP module. The simulator did not provide haptic feedback, but during the tasks the participants received additional visual feedback from color changes indicating success or errors. All participants initially performed at least two repetitions of three different basic tasks, two modules on “camera navigation” and one on “tissue manipulation” in the SimSurgery Educational Platform. Levels of performance were pre-set by the manufacturers and were the same during training and testing. The participants had to obtain a score of >80% twice before advancing to the next task. Before ending the pre-study training, all participants had to successfully perform 10 right-sided salpingectomies in the EP module. The aim was to perform a salpingectomy using the bipolar grasper to cauterize tissue, followed by scissors to cut. Once the ectopic had been fully separated from the Fallopian tube, it was to be placed in a virtual endo-bag and any residual bleeding controlled. The simulation ended when the participant conveyed that the task was completed. Besides “missing removal of specimen,” “failed completion of procedure” and “incomplete hemostasis,” no requirements were set. The VR test consisted of the same EP-module. Participants’ simulated laparoscopic skills were evaluated by three metrics subsequently recorded in the computer. “Time to complete procedure” was the time from the start of the simulation until the end, “tip trajectory” was total length of instrument movements throughout the procedure measured in centimeters, and “blood loss” was the amount of bleeding during the simulated procedure measured in milliliters. To investigate construct validity for these metrics in the EP-module, a minor validation study was done prior to the main study. Three skilled gynecological laparoscopic
ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 1255–1261
Impact of night shifts on gynecologists
A. Veddeng et al.
surgeons were enrolled in the study to serve as a reference group. Each of them was performing more than 100 laparoscopies per year including advanced laparoscopies at level III and IV [European Society for Gynaecological Endoscopy (ESGE) classification]. After being introduced to the EP-module in the VR simulator, they successfully completed 10 right-sided salpingectomies. Their mean performances of the salpingectomy procedure were compared with the mean performances of the participants. After finishing the period of simulator training, the rested or “baseline” test was conducted in a morning between 07.30 and 10.30 h after a normal period of sleep at home. The participants completed three simulated salpingectomies one after another on the VR simulator. Data on the cumulative performance of these three procedures were assessed. Directly before or after the VR test, the participants went through the cognitive testing with the Cambridge neuropsychological test automated battery (CANTAB). The total test time was approximately 45 min. All outcomes were anonymous. Between 2 and 4 months later, the participants went through an identical testing directly after the end of a regular overnight shift from 15.30 to 09.00 h. The time interval between the two test sessions was intended to diminish the potential learning bias in relation to the cognitive testing. Cognitive function was assessed using a selection of cognitive tests from CANTAB (Cambridge Cognition, Cambridge, UK, 2012). The participants were introduced to tests delivered on a touch-sensitive screen and responded by touching the screen and a touch pad. With CANTAB it is possible to report a wide range of results from each test describing different aspects of each function tested. In our study battery we chose one introductory and three cognitive tests that have moderate or good test-retest reliability – Motor Screening Test, Paired Associates Learning, Stocking of Cambridge and Reaction Time (10). Participants completed a questionnaire at the beginning of each test session. Data obtained included position of work, level of experience, gender, age, left-handedness, right-handedness, hours slept, cups of coffee drunk during the preceding 12 h, total number of performed laparoscopies ever, and number of laparoscopies during the last year. These laparoscopies were further specified as diagnostic, therapeutic procedures and advanced procedures. Subjective feeling of sleepiness was registered using the Karolinska Sleepiness Scale (11). The participants were asked to grade their current state of alertness on a scale from 1 to 9, where 1 is extremely alert and 9 is very sleepy. A convenience sample of participants was chosen. The study was conducted in one single department to minimalize deviations regarding length of on-call shifts,
100
5
50
0
50
0
–5 0 –50
–10 –50 –100 –15 Time to perform procedure (s)
Tip trajectory (cm)
Blood loss (ml)
Figure 1. Scatter/dot with boxplot illustrating differences in levels of performance between after on-call and baseline testing. Paired samples t-test.
duty load and hospital routines. All eligible gynecologists were included: 28 gynecologists and three skilled gynecological laparoscopic surgeons. The Statistical Package for the Social Science, version 19.0 (SPSS Inc., Chicago, IL, USA) was used to analyze the data. Differences in mean results between rested and post-call tests were analyzed with paired samples Student t-test (Figure 1). p values
AOGS M A I N R E SE A RC H A R TI C LE
Impact of night shifts on laparoscopic skills and cognitive function among gynecologists ANNE VEDDENG1, THOMAS HUSBY2, INGEBORG B. ENGELSEN1, ANDREW KENT3,4 & HANS FLAATTEN2,5 1
Department of Gynecology and Obstetrics, 2Department of Anesthesiology, Haukeland University Hospital, Bergen, Norway, 3Advanced Gynecological Endoscopy Program, University of Surrey, 4Royal Surrey County Hospital, Guildford, UK and 5Department of Clinical Medicine, University of Bergen, Bergen, Norway
Key words Assessment of surgical skills, virtual reality simulator, sleep deprivation, work-hour restrictions, patient safety Correspondence Anne Veddeng, Department of Gynecology and Obstetrics, Haukeland University Hospital, 5021 Bergen, Norway. E-mail: [email protected] Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. Please cite this article as: Veddeng A, Husby T, Engelsen IB, Kent A, Flaatten H. Impact of night shifts on laparoscopic skills and cognitive function among gynecologists. Acta Obstet Gynecol Scand 2014; 93: 1255–1261. Received: 2 November 2013 Accepted: 25 August 2014 DOI: 10.1111/aogs.12496
Abstract Objective. To assess whether gynecologists have impaired laparoscopic skills and/or reduced cognitive function after long on-call hours. Design. Prospective cohort study. Setting. Department of Gynecology and Obstetrics, Norway. Population. 28 gynecologists working long shifts in a single department. Methods. Pre-training of laparoscopic skills on a virtual reality simulator. Simulator- and cognitive testing on two different occasions; one in the morning after a normal nights’ sleep at home and one in the morning directly after 17.5 h on call. The virtual reality simulator test consisted of three repetitive salpingectomies in an ectopic pregnancy module. The cognitive test consisted of a standardized cognitive test battery (Cambridge neuropsychological test automated battery). Main outcome measures. Simulated laparoscopic performance was assessed by time to perform the procedure, total length of instrument movement (tip trajectory) and blood loss. Cognitive function was assessed by reaction time, errors and total score. Results. No significant impairment in laparoscopic skills was found after 17.5 h on call. Cognitive testing revealed a statistically significant increase in reaction time post-call. Construct validity for the metrics “time to perform procedure” and “tip trajectory” in the ectopic pregnancy module was established in a validation study prior to our main study. Conclusions. We were not able to detect impaired laparoscopic skills among gynecologists tested by a virtual reality procedural module after 17.5 h on call. We found a small increase in reaction time but no other signs of reduced cognitive function. The study adds information on surgical performance of sleep-deprived gynecologists. CANTAB, Cambridge neuropsychological test automated battery; EP, ectopic pregnancy; tip trajectory, total length of instrument movement; VR, virtual reality.
Abbreviations:
Introduction The duration of on-call shifts for medical professionals is a matter of debate and has frequently been the focus of discussion in the Norwegian media as well as of the national labor inspection authority. The media suggest that physicians demand long work hours only to increase
Key Message Simulated laparoscopic skills are not necessarily impaired among professionals in a clinical setting at the end of long periods on call. Cognitive testing may add strength to the assessment of surgical skills among sleep-deprived surgeons.
ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 1255–1261
1255
Impact of night shifts on gynecologists
salary, jeopardizing patient safety (1). Norwegian working hour regulations limit physicians to work a maximum of 19 consecutive hours and a maximum of 60 h a week. The European Working Time Directive for physicians in the European Union is even stricter, with a maximum of 13 consecutive hours and 48 h a week (since August 2009). Due to increased focus on patient safety there is a demand for consultants to be available on-site 24/7 in Norwegian hospitals, not at least in the field of obstetrics and gynecology. Physicians are exempted from their daily work before and after being on call, and shorter duration and more frequent on-calls will therefore lead to a lack of consultants to cover specialized tasks in the daytime. Moreover, physicians’ clinical skills are at risk of being impaired, as these are mainly developed and maintained by daytime work. Research has demonstrated that one night of prolonged wakefulness causes performance impairments equivalent to an alcohol intoxication of 0.10% (2). Although there is little doubt that sleep deprivation impairs human functioning, typical surgical skills do not necessarily deteriorate with a limited amount of sleep loss under average conditions in a hospital or clinic (3,4). For obvious reasons, no randomized controlled trials have examined the direct effects of surgeon fatigue on patient outcome, and therefore simulation studies have frequently been used in this field of research (4–9). However, few trials have included surgeons in real clinical settings. Cognitive testing has often been omitted, although it is widely known that surgery consists of good judgment as well as technical skills. Gynecologists/obstetricians on call are prone to sleep deprivation and they often have to perform emergency surgical procedures during their night shifts. Primarily, we aimed to investigate simulated laparoscopic skills and cognitive function among gynecologists immediately after 17.5 h on call as compared to a rested situation. Secondarily, we wished to investigate the construct validity of relevant metrics in the ectopic pregnancy (EP) module in the SimSurgery virtual reality (VR) simulator.
Material and methods Twenty-eight of 33 gynecologists working on-call shifts at the Department of Gynecology and Obstetrics, Haukeland University Hospital, were enrolled in the study between October and November 2011. The reasons for not participating were planned study or maternity leave in the period of trial. All participants gave informed consent under a protocol approved by the Regional Ethic Committee in Norway (REK Vest, 2011/1900). The trial was completed in April 2012.
1256
A. Veddeng et al.
Haukeland University Hospital is a secondary and tertiary health care facility, providing health services to approximately 1 million inhabitants living in the western part of Norway. The annual number of deliveries at the Department of Gynecology and Obstetrics is approximately 5000. The gynecologists working on-call have on average one on-call shift a week (15.30–09.00 h), and they are exempted from regular working hours before and after being on call. Due to the intensive workload, and for educational and safety reasons, gynecologists from three different tiers, a junior trainee, a more experienced trainee and an experienced specialist, stay at the hospital during the whole on-call shift. A minority of the participants had experience with the VR simulator SimSurgery Educational Platform (SimSurgery AS, Oslo, Norway) before the start of this study, but none during the past year. To familiarize the participants with the test tool and to diminish learning bias during the study period, they received pre-study supervised training on the VR simulator, using the SimSurgery Educational Platform Basic and EP module. The simulator did not provide haptic feedback, but during the tasks the participants received additional visual feedback from color changes indicating success or errors. All participants initially performed at least two repetitions of three different basic tasks, two modules on “camera navigation” and one on “tissue manipulation” in the SimSurgery Educational Platform. Levels of performance were pre-set by the manufacturers and were the same during training and testing. The participants had to obtain a score of >80% twice before advancing to the next task. Before ending the pre-study training, all participants had to successfully perform 10 right-sided salpingectomies in the EP module. The aim was to perform a salpingectomy using the bipolar grasper to cauterize tissue, followed by scissors to cut. Once the ectopic had been fully separated from the Fallopian tube, it was to be placed in a virtual endo-bag and any residual bleeding controlled. The simulation ended when the participant conveyed that the task was completed. Besides “missing removal of specimen,” “failed completion of procedure” and “incomplete hemostasis,” no requirements were set. The VR test consisted of the same EP-module. Participants’ simulated laparoscopic skills were evaluated by three metrics subsequently recorded in the computer. “Time to complete procedure” was the time from the start of the simulation until the end, “tip trajectory” was total length of instrument movements throughout the procedure measured in centimeters, and “blood loss” was the amount of bleeding during the simulated procedure measured in milliliters. To investigate construct validity for these metrics in the EP-module, a minor validation study was done prior to the main study. Three skilled gynecological laparoscopic
ª 2014 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 1255–1261
Impact of night shifts on gynecologists
A. Veddeng et al.
surgeons were enrolled in the study to serve as a reference group. Each of them was performing more than 100 laparoscopies per year including advanced laparoscopies at level III and IV [European Society for Gynaecological Endoscopy (ESGE) classification]. After being introduced to the EP-module in the VR simulator, they successfully completed 10 right-sided salpingectomies. Their mean performances of the salpingectomy procedure were compared with the mean performances of the participants. After finishing the period of simulator training, the rested or “baseline” test was conducted in a morning between 07.30 and 10.30 h after a normal period of sleep at home. The participants completed three simulated salpingectomies one after another on the VR simulator. Data on the cumulative performance of these three procedures were assessed. Directly before or after the VR test, the participants went through the cognitive testing with the Cambridge neuropsychological test automated battery (CANTAB). The total test time was approximately 45 min. All outcomes were anonymous. Between 2 and 4 months later, the participants went through an identical testing directly after the end of a regular overnight shift from 15.30 to 09.00 h. The time interval between the two test sessions was intended to diminish the potential learning bias in relation to the cognitive testing. Cognitive function was assessed using a selection of cognitive tests from CANTAB (Cambridge Cognition, Cambridge, UK, 2012). The participants were introduced to tests delivered on a touch-sensitive screen and responded by touching the screen and a touch pad. With CANTAB it is possible to report a wide range of results from each test describing different aspects of each function tested. In our study battery we chose one introductory and three cognitive tests that have moderate or good test-retest reliability – Motor Screening Test, Paired Associates Learning, Stocking of Cambridge and Reaction Time (10). Participants completed a questionnaire at the beginning of each test session. Data obtained included position of work, level of experience, gender, age, left-handedness, right-handedness, hours slept, cups of coffee drunk during the preceding 12 h, total number of performed laparoscopies ever, and number of laparoscopies during the last year. These laparoscopies were further specified as diagnostic, therapeutic procedures and advanced procedures. Subjective feeling of sleepiness was registered using the Karolinska Sleepiness Scale (11). The participants were asked to grade their current state of alertness on a scale from 1 to 9, where 1 is extremely alert and 9 is very sleepy. A convenience sample of participants was chosen. The study was conducted in one single department to minimalize deviations regarding length of on-call shifts,
100
5
50
0
50
0
–5 0 –50
–10 –50 –100 –15 Time to perform procedure (s)
Tip trajectory (cm)
Blood loss (ml)
Figure 1. Scatter/dot with boxplot illustrating differences in levels of performance between after on-call and baseline testing. Paired samples t-test.
duty load and hospital routines. All eligible gynecologists were included: 28 gynecologists and three skilled gynecological laparoscopic surgeons. The Statistical Package for the Social Science, version 19.0 (SPSS Inc., Chicago, IL, USA) was used to analyze the data. Differences in mean results between rested and post-call tests were analyzed with paired samples Student t-test (Figure 1). p values
