Inr. 3 .%urr. Stud.. Vol. 27. Printed m Great Britain.
No.
3. PP.
291402.
oo?o-7489/90 SJ.oo-0.a) B 1990 Pcrgamon Prcrs plc
1990.
A comparison of rotating-shift and permanent night nurses RUTH R. ALWARD,
Ed.D., R.N.*
President of Nurse Executive Associates Inc. Washington, D. C. U.S.A.
TIMOTHY
H. MONK, Ph.D.
Sleep Evaluation Center, Western Psychiatric Institute and Clinic, University of Pittsburgh School of Medicine, u.s..-l.
Abstract-This study tests the hypothesis that permanent night nurses would be no better off than rotating shift nurses on the first night of a run of duty. Thirty permanent and 30 rotating shift female nurses from the same hospital participated in a study involving sleep diaries, oral temperature measurement, as well as subjective ratings of well-being and the effort needed to complete their work. Although the permanent nurses fared significantly better than rotating shift nurses in subjective ratings, this did not appear to be due to the maintenance of a nocturnal orientation in the permanent group. Significant differences between the groups only emerged in the amount of night sleep before the shift. Day sleep amounts were similar between groups, as was the size of the temperature drop (0S”F) between midnight and 4 a.m. Thus, lifestyle and social differences between the groups appeared to be more likely than biological rhythm adjustment ones to account for the permanent group’s better subjective ratings.
Introduction
Conventional wisdom in the shift work field, at least in the United States, holds that permanent night work is preferable to weekly rotating or irregular night work. Thus, permanent night workers have a greater commitment to a nocturnal routine, promoting a more suitable alignment to their biological clocks (circadian system) and allowing better daytime sleeps and night-time job performance. When shift rotation cannot be avoided, *Author to whom correspondence
should be addressed: 2011 N Street, NW, Washington, DC, 20036, U.S.A. 297
298
RUTH
R. AL WARD
AND
T’IMOTHY
H. MONK
some authors (Czeisler et al., 1982) assert that it should be as slow as possible (e.g. three weeks of night work before changing to a new shift timing) because that represents a compromise that is suitably close to the permanent situation. In contrast, many European shift work authorities favor a very rapid rotation, with one or two nights of work before changing to another shift (Knauth et a/., 1978). Theoretically, biological circadian rhythms of rapidly rotating night workers remain oriented to a day activity/night sleep pattern and experience no adjustment to the brief alteration in schedule. Weekly rotating and irregular shift systems are held by almost every author to be unsatisfactory, because the circadian system is just beginning to achieve a correct orientation for one shift timing (the process takes a week or more) when another shift timing is imposed (Knauth, et al., 1978). In a recent paper, one of us (Monk, 1986) suggested that permanent night work may not be as positive in its benefits as originally thought. Most people are required to live a diurnal routine on their ‘off’ days in order to fulfil social and domestic commitments (Walker, 1985). Moreover, strong time cues, such as daylight/darkness and knowledge of clock time, pull towards a diurnal orientation. Thus, the nocturnal orientation of even permanent night workers’ circadian systems may be lost after a weekend break (van Loon, 1963); and in reality such workers may be rotating, on a weekly basis, between nocturnal and diurnal circadian orientations. From such considerations one might hypothesize that, at the extreme, permanent night workers are no better off than rotating ones at the beginning of a tour of night duty, immediately following an off-duty or weekend-style break. The purpose of this brief report is to test that hypothesis, comparing groups of permanent and rotating shift night nurses. Hypotheses The null hypothesis predicted that no difference would exist between permanent and rotating nurses in: (i) the amount of night (10 p.m.-8 a.m.) sleep taken in the 30 hours before the first night shift, under the alternate hypothesis (H J that permanent night nurses would be expected to be more nocturnally oriented and thus take less night sleep; (ii) the amount of day sleep (8 a.m.-l 1 p.m.) taken in the 30 hours before and after the first night shift (HI: permanent night nurses would take more); (iii) the drop in temperature between midnight and 4 a.m. (H 1: permanent night nurses would show less of a drop or perhaps even a rise reflecting their nocturnal orientation); (iv) well-being at 4 a.m. (HI: permanent night nurses should feel better) and (v) the amount of effort required to complete the shift (H ,: permanent night nurses should find it easier).
Methods
The study was primarily one that compared job performance ratings of permanent and rotating night shift nurses (Alward, 1986; Alward, 1990). The present report focuses on a subsidiary aspect of that research, namely, a comparison of fixed shift and rotating schedule night nurses in measures of sleep timing, oral temperature, and subjective ratings. The study was made at the very beginning of a run of night duty, immediately following an off-duty (weekend-style) break. Sample Two groups of 38 female registered nurses, ages 21-45 years, were recruited from a single 1000 bed metropolitan university hospital. Eighty-eight per cent of those approached who
SHIFT
WORK
299
were eligible for the study agreed to take part when asked. All worked 8.5-hr night shifts (11 p.m.-7.30 a.m.), half on a fixed shift schedule working only nights, and half on a rotating schedule that also included day shifts, changing shift on an irregular basis. The subjects were staff nurses assigned to non-critical care units. The workloads were similar on all units used for the study. Participants had a minimum of three months’ experience on their shift schedules. Although the rotating nurses did not choose to work the night shift, the permanent nurses did make this choice, albeit for different reasons. Forty-two per cent of the latter chose permanent night work for family reasons or were enrolled in advanced education programs. Another 26% of the permanent groups agreed to this shift because it was the only option for a specialty unit assignment (pediatrics, maternity) when they were hired. Permanent night shift selection for the rest of this group had to do with preferring the hours and life style, the differential salary, or the quieter hospital environment at night. Because of the chronic shortage of permanent night nurses, the participating nursing units depended on rotating nurses to fill the night schedules. Selection of two smaller groups of 30 subjects each was then made to balance for average age (permanent: 27.1 y, rotating: 28.1 y), years of nursing experience (permanent: 3.2 y, rotating: 3.1 y), and Horne-Ostberg (1976) Morningness score (permanent: 49.6, rotating: 51.8). The permanent night nurse group differed from the rotating group in the average number of children at home (permanent: 0.47, rotating: 0.23) and in the number who were pursuing nursing degrees at that time (permanent: 9130, rotating: O/30). Procedure
In particular, the analysis was concerned with: (i) sleep diary information from approximately 30 hr either side of the beginning of the first night shift; (ii) measurements of oral temperature at midnight and 4 a.m. on the first night shift, giving a rough indication of circadian phase, because day-oriented individuals typically show a 0.5”F drop between the two (Colquhoun, 1971); (iii) visual analogue scale (VAS) ratings of well-being at 4 a.m. (characteristically, the worst point of the ‘graveyard’ shift) and (iv) ratings of how much effort was required to complete the first night shift (VAS ratings from an ‘end of shift’ questionnaire). Sleep diaries were taken by the subjects and kept by the bed. Questions related to the timing and duration of sleep episodes and naps, and to the method of and mood on awakening from major sleep periods. The present analysis was restricted to information available for both major sleeps and naps, viz. timings of ‘going to bed’ and ‘getting up’. During the night shift, oral temperatures were measured at midnight and 4 a.m. using a Marshall basal temperature (mercury-in-glass) thermometer, calibrated in tenths of a degree F and inserted sublingually for a timed 4 min. Smoking, eating and drinking were prohibited for the I5 min prior to the measurement. Well-being at 4 a.m. was assessed using a VAS with the question ‘How do you feel?’ above a 10 cm line, with ‘very bad’ at the left end and ‘very good’ at the right. Subjects placed a mark somewhere along the line to indicate their feelings at that time. The measure used was the distance in mm of the mark from the left end. Within 30 minutes of the end of the night shift, subjects completed an ‘end of shift’ questionnaire that included a VAS scale of effort (Question: “Overall, how much of an effort was it to carry out your work during this shift?“, anchors: ‘very little’, ‘very much’). Again, the measure taken was the distance of the mark from the left (‘very little’) end in mm.
300
RUTH
R. AL WARD
AND
TIMOTHY
H. MONK
Results Distribution of sleep The percentage of each group of nurses who were trying to sleep as a function of time of day is plotted in Fig. 1. During the last night sleep of the off-duty break, there were indeed differences between permanent and rotating shift nurses, with a greater concentration of time in bed during the night hours (10 p.m. to 8 a.m.) for the rotating group (permanent: 6.63 hr. rotating shift 7.68 hr, t = 2.12, d.f. = 58, p < 0.02). However, the day sleeps (8 a.m. to 11 p.m.) before and after the first night shift failed to show any reliable difference between the groups (before shift: permanent: 3.28 hr, rotating shift: 2.48 hr, t = 1.42, d.f. = 28, n.s.; after shift: permanent: 5.32 hr, rotating shift: 5.16 hr, t < 1, n.s.). Thus, while the null hypothesis regarding night sleeps could be rejected, those regarding day sleeps could not. I
Night 2nd da) 4eeps prior to night shit’t
Day sleep:, following night shift
; rotating shift: - - - - - - - - -_) Fig. 1. The percentage of each group (permanent: who were attempting sleep, plotted as a function of time of day for approximately 30 hr before (upper panel) and after (lower panel) the beginning of the first night shift of a run of duty.
SHIFT
Drop in temperature
WORK
301
between midnight and 4 a.m.
From normative date collected by Colquhoun (1971), one would expect about a 05°F drop in body temperature between midnight and 4 a.m. in normal, diurnally entrained individuals. Interestingly, this was almost exactly the magnitude of the drop in temperature observed in both groups (permanent: 0.5 1, rotating shift: 0.58). A 2-way analysis of variance having time (midnight vs 4 a.m.) as one factor and group (permanent vs rotating shift) as the other failed to show a significant time x group interaction [F(1,58) c 1, n.s.] and the null hypothesis regarding temperature was therefore not rejected. There were, however, significant main effects of both time [F(1,58) = 72.3 1, p < O.OOl]and group [F(1,58) = 5.99, p < 0.0251. Subjective ratings
Both of the null hypotheses regarding subjective ratings could be rejected with confidence. Well-being at 4 a.m. was significantly higher in permanent than in rotating night shift nurses (permanent: 55.5, rotating shift: 43.8, U = 326.5, p < 0.05). Ratings of how much effort was required to complete the shift also showed the permanent group to be better off (permanent: 36.9, rotating shift: 54.2, U = 288.5, p < 0.01).
Discussion
From the present results it would clearly be wrong to conclude that permanent night workers are no better off than rotating shift workers at the beginning of a run of night duty. Permanent night nurses felt better at 4 a.m. on that shift, and perceived the shift as requiring less overall effort than did the rotating shift nurses. However, the explanation for those group differences would not appear to be mediated by simple differences in the amount of nocturnal orientation of the biological clock. Although the permanent group did have less time in bed on the night before the shift, there were no reliable differences between the two groups either in the amount of day sleep taken before or after the shift or in the size of the drop in temperature between midnight and 4 a.m. Indeed, the latter would suggest that even the permanent night nurses were essentially diurnal as far as their endogenous biological clock was concerned [in line with earlier night nurse findings (Monk, 1986; after Folkard, Monk and Lobban, 1978)j. Of direct relevance to the present study is the publication of Verhaegen et al. (1987), which only appeared in print after the present analysis had been completed. Verhaegen et ai.‘s study was also concerned with a comparison of the adaptation of night nurses to different work schedules, using subjective ratings and sleep diary information as well as other variables. Unlike the present study, Verhaegen et al. were concerned with a whole month, rather than the beginning of a run of night duty, and considered morningness (Horne and Ostberg, 1976) as a dependent variable, revealing differences between permanent and rotating schedules, rather than equating morningness score between groups as the present authors have done. The nurses worked in six Belgian hospitals. They were grouped into categories of full-time permanent night nurses, part-time permanent night nurses, and rotating night nurses. Gratifyingly, there appear to be few discrepancies between the conclusion of Verhaegen et al.‘s study and those of our own. Thus, using an Evaluation of Night Work questionnaire, they found better subjective ratings in the full-time permanent night nurses than in the rotators. Although not directly comparable because of the entire month of coverage, their
302
RUTH R. AL WARD AND TIMOTHY
H. MONK
sleep diary data revealed no significant difference between full-time permanent night nurses and rotators in day sleep duration. Although Verhaegen et al. did not take circadian measures, their findings are congruent with our own conclusion that the permanent night nurses did not retain a nocturnal circadian orientation. Most important, Verhaegen et al.‘s investigations regarding the mechanisms by which full-time permanent night nurses fared better than rotating nurses provide empirical support for our own intuitions regarding the present findings. From interview data Verhaegen et al. determined that “very few . . . permanent night nurses seemed to be subject to excessive strain by their work”, perhaps because “permanent night nurses freely chose their working schedule and organized their family life accordingly”. That explanation seems to be a plausible one for explaining the results of the present study. The regularity of permanent night work allows one a good opportunity to build a pattern of living in relationships, child care, and domestic chores that copes to some extent with the unusual work hours even if one’s circadian system cannot retain a nocturnal orientation. For the 42% of permanent night nurses who chose this work schedule to accommodate their families or personal educational goals, the decision was made to establish a more tolerable, less stressful pattern of living. In contrast, the rotating shift worker, particularly when on an irregular rota, often regards the night shift as an occasional nuisance, one that disrupts the pattern of living and has to be endured rather than enjoyed. In attempting to develop the most effective shift schedules and alleviate the problems of shift workers, it behooves researchers and nurse administrators to consider all of the relevant factors, be they social or biological in origin. Concentration on strictly biological factors may lead to erroneous conclusions. Acknowledgemenfs-Thanks data analysis.
are due to the nurses and their management, and to Miss S. M. Peetz for help in
References Alward, R. R. (1990). Performance of night nurses on permanent and rotating shift schedules. (In preparation.) Alward, R. R. (1986). Performance of permanent versus rotating night nurses: Circadian-related factors. Unpublished doctoral dissertation, Teachers College, Columbia University, New York. Colquhoun, W. P. (1971). Circadian variations in mental efficiency. In Biologicul Rhythms and Human Performonce. W. P. Colquhoun (Ed.), pp. 39-107. Academic Press, London. Czeisler, C. A., Moore-Ede, M. C. and Coleman, R. M. (1982). Rotating shift work schedules that disrupt sleep are improved by applying circadian principles. Science 217. 460-463. Folkard, S., Monk, T. H. and Lobban, N. C. (1978). Short and long-term adjustment of circadian rhythms in ‘permanent’ night nurses. Ergonomics 21, 785-799. Horne, .I. A. and Ostberg, 0. (1976). A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. Int. J. Chronobiol. 4, 97-110. Knauth. P., Rutenfranz, J., Herman, G. and Poppl, S. J. (1978). Re-entrainment of physiological functions in experimental shiftwork studies. Ergonomics 21. 775-783. Monk, T. H. (1986). Advantages and disadvantages of rapidly rotating shift schedules-A circadian viewpoint. Human Factors 28, 553-557. van Loon, J. H. (1963). Diurnal body temperature curves in shift workers. Ergonomics 6, 267-272. Verhaegen, P., Cober, R., De Smedt, M., Dirkx, J., Kerstens, J., Ryvers, D., and Van Daele. P. (1987). The adaptation of night nurses to different work schedules. Ergonomics 30, 1301-1309. Walker, J. (1985). Social problems of shift work. In Hours of Work-Temporal Fucfors in Work Scheduling. S. Folkard and T. H. Monk (Eds), pp. 211-225. John Wiley, New York. (Received 14 September 1989; accepted for publication 10 January 1990)
No.
3. PP.
291402.
oo?o-7489/90 SJ.oo-0.a) B 1990 Pcrgamon Prcrs plc
1990.
A comparison of rotating-shift and permanent night nurses RUTH R. ALWARD,
Ed.D., R.N.*
President of Nurse Executive Associates Inc. Washington, D. C. U.S.A.
TIMOTHY
H. MONK, Ph.D.
Sleep Evaluation Center, Western Psychiatric Institute and Clinic, University of Pittsburgh School of Medicine, u.s..-l.
Abstract-This study tests the hypothesis that permanent night nurses would be no better off than rotating shift nurses on the first night of a run of duty. Thirty permanent and 30 rotating shift female nurses from the same hospital participated in a study involving sleep diaries, oral temperature measurement, as well as subjective ratings of well-being and the effort needed to complete their work. Although the permanent nurses fared significantly better than rotating shift nurses in subjective ratings, this did not appear to be due to the maintenance of a nocturnal orientation in the permanent group. Significant differences between the groups only emerged in the amount of night sleep before the shift. Day sleep amounts were similar between groups, as was the size of the temperature drop (0S”F) between midnight and 4 a.m. Thus, lifestyle and social differences between the groups appeared to be more likely than biological rhythm adjustment ones to account for the permanent group’s better subjective ratings.
Introduction
Conventional wisdom in the shift work field, at least in the United States, holds that permanent night work is preferable to weekly rotating or irregular night work. Thus, permanent night workers have a greater commitment to a nocturnal routine, promoting a more suitable alignment to their biological clocks (circadian system) and allowing better daytime sleeps and night-time job performance. When shift rotation cannot be avoided, *Author to whom correspondence
should be addressed: 2011 N Street, NW, Washington, DC, 20036, U.S.A. 297
298
RUTH
R. AL WARD
AND
T’IMOTHY
H. MONK
some authors (Czeisler et al., 1982) assert that it should be as slow as possible (e.g. three weeks of night work before changing to a new shift timing) because that represents a compromise that is suitably close to the permanent situation. In contrast, many European shift work authorities favor a very rapid rotation, with one or two nights of work before changing to another shift (Knauth et a/., 1978). Theoretically, biological circadian rhythms of rapidly rotating night workers remain oriented to a day activity/night sleep pattern and experience no adjustment to the brief alteration in schedule. Weekly rotating and irregular shift systems are held by almost every author to be unsatisfactory, because the circadian system is just beginning to achieve a correct orientation for one shift timing (the process takes a week or more) when another shift timing is imposed (Knauth, et al., 1978). In a recent paper, one of us (Monk, 1986) suggested that permanent night work may not be as positive in its benefits as originally thought. Most people are required to live a diurnal routine on their ‘off’ days in order to fulfil social and domestic commitments (Walker, 1985). Moreover, strong time cues, such as daylight/darkness and knowledge of clock time, pull towards a diurnal orientation. Thus, the nocturnal orientation of even permanent night workers’ circadian systems may be lost after a weekend break (van Loon, 1963); and in reality such workers may be rotating, on a weekly basis, between nocturnal and diurnal circadian orientations. From such considerations one might hypothesize that, at the extreme, permanent night workers are no better off than rotating ones at the beginning of a tour of night duty, immediately following an off-duty or weekend-style break. The purpose of this brief report is to test that hypothesis, comparing groups of permanent and rotating shift night nurses. Hypotheses The null hypothesis predicted that no difference would exist between permanent and rotating nurses in: (i) the amount of night (10 p.m.-8 a.m.) sleep taken in the 30 hours before the first night shift, under the alternate hypothesis (H J that permanent night nurses would be expected to be more nocturnally oriented and thus take less night sleep; (ii) the amount of day sleep (8 a.m.-l 1 p.m.) taken in the 30 hours before and after the first night shift (HI: permanent night nurses would take more); (iii) the drop in temperature between midnight and 4 a.m. (H 1: permanent night nurses would show less of a drop or perhaps even a rise reflecting their nocturnal orientation); (iv) well-being at 4 a.m. (HI: permanent night nurses should feel better) and (v) the amount of effort required to complete the shift (H ,: permanent night nurses should find it easier).
Methods
The study was primarily one that compared job performance ratings of permanent and rotating night shift nurses (Alward, 1986; Alward, 1990). The present report focuses on a subsidiary aspect of that research, namely, a comparison of fixed shift and rotating schedule night nurses in measures of sleep timing, oral temperature, and subjective ratings. The study was made at the very beginning of a run of night duty, immediately following an off-duty (weekend-style) break. Sample Two groups of 38 female registered nurses, ages 21-45 years, were recruited from a single 1000 bed metropolitan university hospital. Eighty-eight per cent of those approached who
SHIFT
WORK
299
were eligible for the study agreed to take part when asked. All worked 8.5-hr night shifts (11 p.m.-7.30 a.m.), half on a fixed shift schedule working only nights, and half on a rotating schedule that also included day shifts, changing shift on an irregular basis. The subjects were staff nurses assigned to non-critical care units. The workloads were similar on all units used for the study. Participants had a minimum of three months’ experience on their shift schedules. Although the rotating nurses did not choose to work the night shift, the permanent nurses did make this choice, albeit for different reasons. Forty-two per cent of the latter chose permanent night work for family reasons or were enrolled in advanced education programs. Another 26% of the permanent groups agreed to this shift because it was the only option for a specialty unit assignment (pediatrics, maternity) when they were hired. Permanent night shift selection for the rest of this group had to do with preferring the hours and life style, the differential salary, or the quieter hospital environment at night. Because of the chronic shortage of permanent night nurses, the participating nursing units depended on rotating nurses to fill the night schedules. Selection of two smaller groups of 30 subjects each was then made to balance for average age (permanent: 27.1 y, rotating: 28.1 y), years of nursing experience (permanent: 3.2 y, rotating: 3.1 y), and Horne-Ostberg (1976) Morningness score (permanent: 49.6, rotating: 51.8). The permanent night nurse group differed from the rotating group in the average number of children at home (permanent: 0.47, rotating: 0.23) and in the number who were pursuing nursing degrees at that time (permanent: 9130, rotating: O/30). Procedure
In particular, the analysis was concerned with: (i) sleep diary information from approximately 30 hr either side of the beginning of the first night shift; (ii) measurements of oral temperature at midnight and 4 a.m. on the first night shift, giving a rough indication of circadian phase, because day-oriented individuals typically show a 0.5”F drop between the two (Colquhoun, 1971); (iii) visual analogue scale (VAS) ratings of well-being at 4 a.m. (characteristically, the worst point of the ‘graveyard’ shift) and (iv) ratings of how much effort was required to complete the first night shift (VAS ratings from an ‘end of shift’ questionnaire). Sleep diaries were taken by the subjects and kept by the bed. Questions related to the timing and duration of sleep episodes and naps, and to the method of and mood on awakening from major sleep periods. The present analysis was restricted to information available for both major sleeps and naps, viz. timings of ‘going to bed’ and ‘getting up’. During the night shift, oral temperatures were measured at midnight and 4 a.m. using a Marshall basal temperature (mercury-in-glass) thermometer, calibrated in tenths of a degree F and inserted sublingually for a timed 4 min. Smoking, eating and drinking were prohibited for the I5 min prior to the measurement. Well-being at 4 a.m. was assessed using a VAS with the question ‘How do you feel?’ above a 10 cm line, with ‘very bad’ at the left end and ‘very good’ at the right. Subjects placed a mark somewhere along the line to indicate their feelings at that time. The measure used was the distance in mm of the mark from the left end. Within 30 minutes of the end of the night shift, subjects completed an ‘end of shift’ questionnaire that included a VAS scale of effort (Question: “Overall, how much of an effort was it to carry out your work during this shift?“, anchors: ‘very little’, ‘very much’). Again, the measure taken was the distance of the mark from the left (‘very little’) end in mm.
300
RUTH
R. AL WARD
AND
TIMOTHY
H. MONK
Results Distribution of sleep The percentage of each group of nurses who were trying to sleep as a function of time of day is plotted in Fig. 1. During the last night sleep of the off-duty break, there were indeed differences between permanent and rotating shift nurses, with a greater concentration of time in bed during the night hours (10 p.m. to 8 a.m.) for the rotating group (permanent: 6.63 hr. rotating shift 7.68 hr, t = 2.12, d.f. = 58, p < 0.02). However, the day sleeps (8 a.m. to 11 p.m.) before and after the first night shift failed to show any reliable difference between the groups (before shift: permanent: 3.28 hr, rotating shift: 2.48 hr, t = 1.42, d.f. = 28, n.s.; after shift: permanent: 5.32 hr, rotating shift: 5.16 hr, t < 1, n.s.). Thus, while the null hypothesis regarding night sleeps could be rejected, those regarding day sleeps could not. I
Night 2nd da) 4eeps prior to night shit’t
Day sleep:, following night shift
; rotating shift: - - - - - - - - -_) Fig. 1. The percentage of each group (permanent: who were attempting sleep, plotted as a function of time of day for approximately 30 hr before (upper panel) and after (lower panel) the beginning of the first night shift of a run of duty.
SHIFT
Drop in temperature
WORK
301
between midnight and 4 a.m.
From normative date collected by Colquhoun (1971), one would expect about a 05°F drop in body temperature between midnight and 4 a.m. in normal, diurnally entrained individuals. Interestingly, this was almost exactly the magnitude of the drop in temperature observed in both groups (permanent: 0.5 1, rotating shift: 0.58). A 2-way analysis of variance having time (midnight vs 4 a.m.) as one factor and group (permanent vs rotating shift) as the other failed to show a significant time x group interaction [F(1,58) c 1, n.s.] and the null hypothesis regarding temperature was therefore not rejected. There were, however, significant main effects of both time [F(1,58) = 72.3 1, p < O.OOl]and group [F(1,58) = 5.99, p < 0.0251. Subjective ratings
Both of the null hypotheses regarding subjective ratings could be rejected with confidence. Well-being at 4 a.m. was significantly higher in permanent than in rotating night shift nurses (permanent: 55.5, rotating shift: 43.8, U = 326.5, p < 0.05). Ratings of how much effort was required to complete the shift also showed the permanent group to be better off (permanent: 36.9, rotating shift: 54.2, U = 288.5, p < 0.01).
Discussion
From the present results it would clearly be wrong to conclude that permanent night workers are no better off than rotating shift workers at the beginning of a run of night duty. Permanent night nurses felt better at 4 a.m. on that shift, and perceived the shift as requiring less overall effort than did the rotating shift nurses. However, the explanation for those group differences would not appear to be mediated by simple differences in the amount of nocturnal orientation of the biological clock. Although the permanent group did have less time in bed on the night before the shift, there were no reliable differences between the two groups either in the amount of day sleep taken before or after the shift or in the size of the drop in temperature between midnight and 4 a.m. Indeed, the latter would suggest that even the permanent night nurses were essentially diurnal as far as their endogenous biological clock was concerned [in line with earlier night nurse findings (Monk, 1986; after Folkard, Monk and Lobban, 1978)j. Of direct relevance to the present study is the publication of Verhaegen et al. (1987), which only appeared in print after the present analysis had been completed. Verhaegen et ai.‘s study was also concerned with a comparison of the adaptation of night nurses to different work schedules, using subjective ratings and sleep diary information as well as other variables. Unlike the present study, Verhaegen et al. were concerned with a whole month, rather than the beginning of a run of night duty, and considered morningness (Horne and Ostberg, 1976) as a dependent variable, revealing differences between permanent and rotating schedules, rather than equating morningness score between groups as the present authors have done. The nurses worked in six Belgian hospitals. They were grouped into categories of full-time permanent night nurses, part-time permanent night nurses, and rotating night nurses. Gratifyingly, there appear to be few discrepancies between the conclusion of Verhaegen et al.‘s study and those of our own. Thus, using an Evaluation of Night Work questionnaire, they found better subjective ratings in the full-time permanent night nurses than in the rotators. Although not directly comparable because of the entire month of coverage, their
302
RUTH R. AL WARD AND TIMOTHY
H. MONK
sleep diary data revealed no significant difference between full-time permanent night nurses and rotators in day sleep duration. Although Verhaegen et al. did not take circadian measures, their findings are congruent with our own conclusion that the permanent night nurses did not retain a nocturnal circadian orientation. Most important, Verhaegen et al.‘s investigations regarding the mechanisms by which full-time permanent night nurses fared better than rotating nurses provide empirical support for our own intuitions regarding the present findings. From interview data Verhaegen et al. determined that “very few . . . permanent night nurses seemed to be subject to excessive strain by their work”, perhaps because “permanent night nurses freely chose their working schedule and organized their family life accordingly”. That explanation seems to be a plausible one for explaining the results of the present study. The regularity of permanent night work allows one a good opportunity to build a pattern of living in relationships, child care, and domestic chores that copes to some extent with the unusual work hours even if one’s circadian system cannot retain a nocturnal orientation. For the 42% of permanent night nurses who chose this work schedule to accommodate their families or personal educational goals, the decision was made to establish a more tolerable, less stressful pattern of living. In contrast, the rotating shift worker, particularly when on an irregular rota, often regards the night shift as an occasional nuisance, one that disrupts the pattern of living and has to be endured rather than enjoyed. In attempting to develop the most effective shift schedules and alleviate the problems of shift workers, it behooves researchers and nurse administrators to consider all of the relevant factors, be they social or biological in origin. Concentration on strictly biological factors may lead to erroneous conclusions. Acknowledgemenfs-Thanks data analysis.
are due to the nurses and their management, and to Miss S. M. Peetz for help in
References Alward, R. R. (1990). Performance of night nurses on permanent and rotating shift schedules. (In preparation.) Alward, R. R. (1986). Performance of permanent versus rotating night nurses: Circadian-related factors. Unpublished doctoral dissertation, Teachers College, Columbia University, New York. Colquhoun, W. P. (1971). Circadian variations in mental efficiency. In Biologicul Rhythms and Human Performonce. W. P. Colquhoun (Ed.), pp. 39-107. Academic Press, London. Czeisler, C. A., Moore-Ede, M. C. and Coleman, R. M. (1982). Rotating shift work schedules that disrupt sleep are improved by applying circadian principles. Science 217. 460-463. Folkard, S., Monk, T. H. and Lobban, N. C. (1978). Short and long-term adjustment of circadian rhythms in ‘permanent’ night nurses. Ergonomics 21, 785-799. Horne, .I. A. and Ostberg, 0. (1976). A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. Int. J. Chronobiol. 4, 97-110. Knauth. P., Rutenfranz, J., Herman, G. and Poppl, S. J. (1978). Re-entrainment of physiological functions in experimental shiftwork studies. Ergonomics 21. 775-783. Monk, T. H. (1986). Advantages and disadvantages of rapidly rotating shift schedules-A circadian viewpoint. Human Factors 28, 553-557. van Loon, J. H. (1963). Diurnal body temperature curves in shift workers. Ergonomics 6, 267-272. Verhaegen, P., Cober, R., De Smedt, M., Dirkx, J., Kerstens, J., Ryvers, D., and Van Daele. P. (1987). The adaptation of night nurses to different work schedules. Ergonomics 30, 1301-1309. Walker, J. (1985). Social problems of shift work. In Hours of Work-Temporal Fucfors in Work Scheduling. S. Folkard and T. H. Monk (Eds), pp. 211-225. John Wiley, New York. (Received 14 September 1989; accepted for publication 10 January 1990)
