Pain is an evolutionary signal of threat that urges us to act in order to escape from or avoid danger , . A common response to pain is to suspend ongoing activities for some time, whilst intending to resume and complete them later . Although this response may facilitate pain management, it can also have negative consequences for the activity.
The events that take place when an activity is interrupted by pain are described in a theoretical model  informed by the interruption literature outside the field of pain (such as the Memory for Goals model , and the stage model of Boehm-Davis and Remington ). This literature shows that (pain-irrelevant) interruptions mostly disrupt performance, (e.g. , , , ). Activity interruptions by pain have also been shown to impair quality of performance, though not more so than interruptions by non-painful stimuli , . Existing evidence indicates that activity interruptions by pain may also influence the way that people perform an activity. For example, people with pain complaints report continuing work outside working hours on days that they experience activity interruptions by pain . Further, healthy people scoring high in pain catastrophizing are shown to persist less in an activity that is interrupted by pain, compared to when it is not interrupted at all . Whether activity interruptions by pain influence the pattern of activity performance differently than activity interruptions by non-pain, however, is unknown.
Whereas the natural tendency is to avoid pain and pain-related activities , , especially when these are not highly valued , at the same time, interrupted activities urge their completion . As a result, people might shift between approaching and avoiding the activity. When the activity cannot be completely avoided (e.g. by being delegated to another person), this may be expressed as prolonged time away from the activity as a temporary solution to the conflict or until pain subsides . The same is not expected for interruptions by non-painful and non-aversive stimuli, in which completing the (unavoidable) task is expected to be prioritized over the interruption. Activities that are interspersed with long interruptions are expected to also need long time for their completion, unless compensatory strategies are used to counteract this negative effect , . Finally, given that the threat value of pain enhances its disruptiveness (e.g. , , ), effects are expected to be more prominent amongst people who experience pain as threatening, such as high pain catastrophizers.
This experiment investigated how activity interruptions by pain influence the pattern of activity performance. Healthy participants performed an ongoing task, during which they occasionally received either painful or non-painful and non-aversive interruption cues, followed by ongoing task suspension and initiation of an open-ended interruption task. We expected that participants interrupted by pain would spend longer time on the interruption task, and therefore on the activity as a whole. Alternatively, we considered that longer time away from the interrupted task would not be reflected on longer total time on task, because of engagement in compensatory actions. Finally, people interrupted by pain were expected to report lower motivation to perform the interrupted task. Effects were expected to be stronger amongst high pain catastrophizers.
Sixty-three healthy volunteers participated in the experiment. Three participants were excluded from the analyses because of technical problems during the lab session, thus leaving our final sample with 60 participants. Exclusion criteria were checked by means of self-report and comprised pregnancy, past or present diagnosis of a psychiatric and/or neurological disorder, current use of anxiolytic and/or antidepressant medication, cardiovascular disease, acute or chronic pain, presence of an electronic implant (e.g. cardiac pacemaker), imperfect command of the Dutch language and impaired (uncorrected) eyesight. Participants were University of Leuven students or employees, who participated in the study on an informed consent basis and who received monetary compensation or partial course credit for their participation.
2.2 Interruption cues
Painful stimuli. Electrocutaneous stimuli (square-wave; duration: 700 ms; frequency: 10 Hz) served as the interruption cues for the pain group. The electrocutaneous stimuli were generated by a DS5 constant current stimulator (Digitimer Limited, Hertfordshire, UK) and delivered through two 8 mm Ag/AgCl surface electrodes (Bilaney, Düsseldorf, Germany). The electrodes were filled with electroconductive gel (K-Y gel, Johnson & Johnson, New Brunswick, NJ, USA) and applied on the dorsal side of the participant’s non-dominant wrist with an inter-electrode distance of ~1 cm. Before electrode attachment, the participant’s skin was scrubbed with a commercially available scrub cream to reduce skin resistance.
Stimulus intensity was individually determined in such a way that the stimuli required some effort to tolerate. After electrode application, the experimenter administered a series of electrocutaneous stimuli, starting with an intensity of 0.2 mA and increasing in steps of 0.4 mA. This procedure continued until the participant did not wish to be administered a higher stimulus, or until they had rated the last stimulus as a 9 on an 11-point “effort to tolerate” scale (0=“no effort at all”; 10=“maximum effort I can exert”). Several times throughout the stimulus series (including upon the last stimulus), participants were also asked to rate the painfulness and unpleasantness of the stimulus on 11-point numerical scales (0=“not at all painful/unpleasant”; 10=“the most painful/unpleasant that I can imagine”). The average stimulus intensity used in this study was 3.8 mA (SD=1.7, range 0.8–8.6).
Non-painful stimuli. Auditory stimuli (doorbell sound; duration: 700 ms; frequency: 16,000 Hz) served as the interruption cues for the non-pain group. The auditory stimuli were presented binaurally by means of a standard headset (Philips, Amsterdam, The Netherlands).
Stimulus intensity was individually determined in such a way that the auditory stimulus was easily perceivable, but not aversively loud. The participant wore the headset and was presented with the auditory stimulus several times, with the volume adjusted every time until the participant stated that the sound was easy to perceive but not too loud. To match the procedure with that of the pain group, participants also rated the effort to tolerate, the intensity and the unpleasantness of the sound on three 11-point numerical scales similar to the ones used for the pain group.
2.3 Experimental task
A joystick paradigm (based on ) was developed with the aim to simulate a simple motoric task with a clearly defined goal, the pursuit of which is occasionally interrupted in response to pain. The specific task components were as follows (see also Fig. 1):
Ongoing task. Participants moved a joystick in four directions (to the left, to the right, to the screen and to themselves). Each movement corresponded to one of four locations (left, right, up and down) on the computer screen. On each location there were four colored (two blue, two yellow) rectangles. The screen background was grey (Fig. 1A). At the central point between the four locations, a fixation cross was presented, indicating when participants were required to move the joystick to one of the locations. When the movement was registered, the fixation cross disappeared and the first colored rectangle at that location became black and thus unavailable. After an intertrial interval of 2,500 ms, the next rectangle at that location became available. Each participant was assigned a target color (blue or yellow, counterbalanced) and was instructed to prioritize movements to locations with available rectangles of that color. For example, participants with blue as their target color were required to first “free up” and reach as many blue rectangles as they could, before making movements to locations with available yellow rectangles. A vertical bar was presented on the left side of the screen and indicated the participant’s progress in the ongoing task. After every movement towards a target color rectangle, the bar was filled with the target color (Fig. 1B) in such a pace, that it was completely filled only when the participant had made movements to all target color rectangles of the whole task. The task consisted of 20 blocks of 16 movements each. When all rectangles of a block had become unavailable, a new block began immediately, thus giving participants a sense of continuity during the task.
On 10 randomly preselected trials throughout the task, an interruption cue (painful electrocutaneous or non-painful non-aversive auditory stimulus) was administered. It was administered during the joystick movement, with its onset being when the (invisible) joystick cursor left the fixation cross. Upon completion of the movement and the interruption cue, the ongoing task was suspended and the interruption task was initiated.
Interruption task. The interruption task aimed at engaging participants in a similar way during the interruption, and consisted of a simple variation of the Wisconsin card sorting task . In each trial, three cards were presented on a black background; one card was always presented at the center of the screen, whereas the other two cards were presented either on its left and right (horizontal version; 50% of the interruptions), or above and below it (vertical version; 50% of interruptions) (Fig. 1D). Each card depicted one to four identical geometrical shapes, in one of four different colors (e.g. three green stars). There were five different types of shapes. Participants were instructed to move the joystick towards the card that depicted the same geometrical shapes as the middle card. This categorization rule was known to participants and did not change throughout the task. The cards remained on the screen until a response had been given, or for 7,000 ms. Intertrial interval was 1,500 ms.
Each interruption continued for a maximum of 2 min, but participants could also stop it earlier by pressing a Stop-button on the joystick. After pressing the Stop-button or after the 2 min had passed, the interruption task stopped and the ongoing task started again with the next trial and the same screen configuration (i.e. the same available rectangles and status of vertical bar) as it was when interrupted (Fig. 1E). Participants were instructed to perform both the ongoing as well as the interruption task as fast and as accurately as possible.
2.4.1 Behavioral measures
Our main outcomes were (1) the time that participants spent on the interruption task (i.e. time spent away from the interrupted ongoing task), and (2) the total time taken to complete the experimental task (i.e. the time from the first to the last trial of the ongoing task, including the time spent on the 10 interruptions). Additionally, we also measured (3) the number of times that participants pressed the Stop-button to end the interruption.
2.4.2 Self-report measures
In order to investigate the role of pain catastrophizing, we administered the Pain Catastrophizing Scale (PCS); , . The PCS measures rumination about pain, magnification of pain sensations and helplessness about the pain. Participants are asked to consider previous painful experiences and to rate each of 13 items on a 0 (“not at all”) to 4 (“all the time”) scale. The Dutch version of the PCS has previously shown good psychometric qualities in non-clinical samples  and had a Cronbach’s α=0.91 in our sample (n=62).
Participants rated their motivation to perform the ongoing and the interruption task, namely the degree to which they would be willing to perform each of the two tasks again, on two 11-point numerical scales (0=“not at all”, 10=“to a very high degree”).
For our manipulation check, participants rated the interruption cue characteristics, specifically the painfulness (for the pain group only), unpleasantness, and threat value of the interruption cue on 11-point numerical scales (0=“not at all”, 10=“to a very high degree”).
To explore the effects of our experimental manipulations on state affect, we administered the Self-Assessment Manikin (SAM); , and the state version of the Positive and Negative Affect Schedule (PANAS); , . The SAM version we used consisted of three 9-point pictorial scales, each of which measures one of the basic affective dimensions, namely pleasure, arousal and dominance. For the PANAS, participants read 10 adjectives describing positive affect and 10 adjectives describing negative affect, and were required to rate the degree to which each one of them described how they felt at that moment on a 5-point scale.
Other measures 1 were administered for exploratory reasons and will thus not be further discussed.
The study consisted of a lab session and an online questionnaire session that took place 1 week later. To protect integrity, the experiment was presented as an investigation of visuospatial skills. The study protocol was approved by the Ethics Committee of the Department of Psychology of the University of Leuven (reg. nr. S54157) and the Medical Ethics Committee of the University Hospital Leuven (reg. nr. ML8125). During the lab session, participants were tested individually in a dimly lit room, as follows:
Introduction. Upon arrival to the lab, participants were allocated to one of the two interruption cue conditions (painful vs. non-painful) and one of the two target colors (blue vs. yellow) according to a blocked randomization procedure, and read information about the condition they belonged to. Exclusion criteria were checked and participants provided informed consent, demographic information, and the first state affect measurement (SAM and PANAS).
Interruption cue calibration/familiarization. The experimenter applied the electrodes on the wrist of the participant’s non-dominant arm (pain group) or asked the participant to wear the headset (non-pain group), and the stimulus intensity to be used during the task was individually determined (see Section 2.2). Participants filled in the second state affect (SAM only, due to the length of the PANAS).
Experimental task. Participants received oral and written task instructions. First, they read instructions about the ongoing task and practiced one block of 16 movements. Subsequently, they read instructions about the interruption task and the use of the Stop-button, and practiced a maximum of 16 trials. After the fourth trial, participants could also practice using the Stop-button. The practice phase included visual accuracy feedback, trials and ITIs longer than in the test phase, and no interruption cues. Participants were informed that, in the test phase, interruption cues would be administered during the ongoing task but never during the interruption task. Participants were told that they were required to perform a certain number of trials for the ongoing task, as they would be “finished with the task when the [vertical] bar was completely filled [with color]”. They were also told that “every time [the interruption task started] they could perform [it] for a maximum of 2 min” but they could actually “determine [themselves] how many cards to categorize and when to stop [the interruption task]” by pressing the Stop-button. Note that the ongoing and interruption task were presented as the “color” and the “card” task, respectively, in order to not be charged with differential motivational value or importance. It was assumed, however, that the goal of filling the bar with color would motivate participants to resume and complete the ongoing task. After the instruction and practice phase, participants performed 20 blocks of the ongoing task, 10 of which included an interruption cue followed by a maximum of 2 min on the interruption task. To keep the procedure as similar as possible across the two groups, pain group participants were also asked to wear the headset while performing the computer task, allegedly to block external background noise. However, pain group participants received no auditory stimulation.
End session. Upon task completion, the experimenter removed the electrodes and/or headset. Then, participants rated the characteristics of the interruption cue, their motivation to perform each task, and their state affect (SAM and PANAS), and chose their compensation. Participants were debriefed about the real aims and hypotheses of the study when the whole sample had been tested.
The computer task was programmed with Affect, version 4.0  and run on a Windows XP computer (DELL Optiplex 755) with an Intel Core 2 Duo (2.33 GHz) processor. Participants viewed the task on a DELL 19-inch screen with a 1024×768 pixel resolution, and performed it by means of an Attack™ 3 Joystick (Logitech, Lausanne, Switzerland). The online platform Limesurvey  was used for the questionnaire session.
2.7 Statistical analyses
To compare the two groups with regards to their characteristics and the ratings of the interruption cue, we performed a series of χ2-tests and unpaired sample t-tests with interruption cue group (2: pain vs. non-pain) as the independent variable.
For our main hypotheses, we expected to see longer time on the interruption task, longer total time to complete the experimental task and lower self-reported motivation to perform the ongoing task in participants who received the painful interruption cue, compared to participants who received the non-painful and non-aversive interruption cue. To investigate these expectations, we performed the following analyses. First, a mixed analysis of variance (ANOVA) with interruption cue group (2: pain vs. non-pain) as the between-subjects factor and interruption number (10: interruption 1–interruption 10) as a within-subjects factor, on the time on the interruption task. Second, a one-way ANOVA with interruption cue group (2: pain vs. non-pain) on the total time to complete the experimental task. Third, a mixed ANOVA with interruption cue group (2: pain vs. non-pain) as the between-subjects factor and task type (2: ongoing task vs. interruption task) as the within-subjects factor on the self-reported motivation. In addition, we performed a one way ANOVA with interruption cue group (2: pain vs. non-pain) on the number of Stop-Button presses, to explore whether groups would also differ in this respect. In order to investigate the role of pain catastrophizing in the examined relationships, we included the centered PCS score as a continuous variable in the above analyses. These analyses yielded no essentially changed or additional statistically significant results, and are therefore omitted from the main manuscript in the interest of brevity, but are presented in Supplementary Material.
In addition, in order to explore whether and how state affect changed throughout the session, we subjected the SAM and PANAS ratings to two separate mixed ANOVAs with interruption cue group (2: pain vs. non-pain) as the between-subjects factor and time point (for SAM, 3: beginning session, after calibration, end session; for PANAS, 2: beginning session, end session) as the within-subjects factor.
Due to violations of the sphericity assumption, multivariate test statistics (Pillai’s Trace) are reported , , . Where relevant, mean differences with their 95% confidence intervals (CI) are reported. The reported effect size is partial eta squared (ηp2) with its 90% CI . Missing values (one participant did not fill in the online questionnaires) were excluded listwise. Four participants were excluded from the analyses on the total time to complete the experimental task, because of a programming error that prevented the accurate registration of the total duration, thus leaving the sample for these analyses with 56 participants. This error had no impact on registration of other data. Analyses were performed with SPSS version 22.0 . Note that due to the novelty of the study we lacked prior data to base a thorough a priori power analysis on. However, we performed an ad hoc power analysis with G*Power 220.127.116.11 . This analysis indicated that our study was adequately powered to detect a large effect size (required n=30), but not necessarily a medium effect size (required n=70) (calculations made for Mixed ANOVA with two groups and 10 measurements correlated at 0.5, α of 0.05 and a power at the conventional level of 0.80).
3.1 Group characteristics
3.2 Interruption cue characteristics
3.3 Interruption effects
3.3.1 Time on the interruption task
Figure 2 depicts the average time spent on the interruption task across the experimental session, per group. The time spent on the interruption task decreased over time (main effect interruption number: F(9, 50)=3.97, p=0.001, ηp2=0.416 [0.135, 0.469]), but this did not depend on the interruption cue group (main effect: F(1, 58)=0.64, p=0.429, ηp2=0.11 [0, 0.09]; interruption cue group*interruption number: F(9, 50)=1.89, p=0.075, ηp2=0.254 [0, 0.296]). Exactly because the interruption cue group did not have a statistically significant effect, we do not interpret the seeming group difference in the first interruption.
3.3.2 Total time to complete the experimental task
The average total time to complete the experimental task (i.e. from the first to the last ongoing task trial including the 10 times the participants performed the interruption task) was 1,627 s (SD=350, range 1,210–2,406) for the pain group, and 1,689 s (SD=337, range 1,228–2,404) for the non-pain group. The difference between the two groups was not statistically significant, F(1, 54)=0.45, p=0.504, ηp2=0.008 [0, 0.086].
3.3.3 Self-reported motivation
Participants reported a low to moderate motivation to perform again the ongoing task (pain group: mean 3.76, SD=2.63, range 0–8; non-pain group: mean 3.87, SD=2.20, range 0–9) and the interruption task (pain group: mean 5.03, SD=2.81, range 0–9; non-pain group: mean 4.39, SD=2.14, range 0–9). Motivation to perform again the interruption task was higher than that for the ongoing task (main effect task type: F(1, 58)=9.51, p=0.003, ηp2=0.141 [0.030, 0.276]; mean difference 0.90, 95% CI [0.31, 1.48]), but this did not differ for the two interruption cue groups (main effect: F(1, 58)=0.23, p=0.636, ηp2=0.004 [0, 0.067]; interruption cue group*task type, F(1, 58)=1.71, p=0.196, ηp2=0.029 [0, 0.127]).
3.3.4 Stop-button presses during the interruption task
We also explored for potential differences in the number of times that participants stopped the interruptions themselves (possible maximum was 10 times), instead of allowing them to continue to the full 2 min. On average, participants pressed the Stop-button 8.93 (SD=2.62, range: 0–10) times in the pain group and 8.58 (SD=2.62, range: 0–10) times in the non-pain group. This difference did not reach statistical significance, F(1, 58)=0.27, p=0.606, ηp2=0.005 [0, 0.070]. The vast majority of participants (n=40, 66.7%; 20 of the pain group, 20 of the non-pain group) pressed the Stop-button all 10 times, whereas only few participants (n=3, 5%; two of the pain group, one of the non-pain group) never pressed the Stop-button.
3.4 Self-reported state affect
We also explored whether and how state affect changed throughout the session (see Table 3). Below we discuss each (sub)scale separately.
SAM pleasure. In general, participants reported moderate to high levels of pleasure. A main effect of time point, F(2, 55)=12.04, p<0.001, ηp2=0.305 [0.129, 0.430], was qualified by a significant interruption cue group*time point interaction, F(2, 55)=6.98, p=0.002, ηp2=0.202 [0.051, 0.331]. The pain group participants reported lower pleasure than the non-pain group participants after the calibration, p=0.026 (mean difference −0.69, 95% CI [−1.29, −0.09]), but not at the beginning, p=0.69, or end of the session, p=0.729. There was no main effect of interruption cue group, F(1, 56)=0.83, p=0.368, ηp2=0.015 [0, 0.101].
SAM arousal. The reported level of arousal was low. Analyses yielded a significant effect of time point, F(2, 55)=7.3, p=0.002, ηp2=0.21 [0.056, 0.339], superseded by a significant interruption cue group*time point interaction, F(2, 55)=10.52, p<0.001, ηp2=0.277 [0.106, 0.404]. Pain group participants reported significantly higher arousal than non-pain group participants after the calibration, p=0.012 (mean difference 1.16, 95% CI [0.27, 2.06]), but not at the beginning, p=0.754, or end of the session, p=0.071. There was no effect of interruption cue group, F(1, 56)=3.03, p=0.087, ηp2=0.051 [0, 0.166].
SAM dominance. Participants reported a moderate level of dominance. There were no statistically significant effects (main effect interruption cue group: F(1, 56)=2.08, p=0.155, ηp2=0.036 [0, 0.142]; main effect time point: F(2, 55)=0.59, p=0.559, ηp2=0.021 [0, 0.091]; interruption cue group*time point: F(2, 55)=2.69, p=0.077, ηp2=0.089 [0, 0.202]).
PANAS positive. The reported levels of positive affect were moderate, and appeared to decrease throughout the session, F(1, 58)=35.24, p<0.001, ηp2=0.378 [0.214, 0.503] (mean difference −3.57, 95% CI [−4.77, −2.37]). This did not depend on the interruption cue group (main effect: F(1, 58)=1.69, p=0.199, ηp2=0.028 [0, 0.127]; interruption cue group*time: F(1, 58)=0.6, p=0.443, ηp2=0.01 [0, 0.089]).
PANAS negative. Participants reported rather low levels of negative affect. A main effect of time point, F(1, 58)=7.68, p=0.008, ηp2=0.117 [0.018, 0.249], was superseded by an interruption cue group*time point interaction, F(1, 58)=7.07, p=0.01, ηp2=0.109 [0.015, 0.239]. Specifically, negative affect decreased across the session for the non-pain group, p<0.001 (mean difference −1.68, 95% CI [−2.54, −0.82]), but not for the pain group, p=0.938 (mean difference −0.03, 95% CI [−0.92, 0.85]). There was no main effect of interruption cue group, F(1, 58)=0.59, p=0.445, ηp2=0.01 [0, 0.089].
Taken together, our results indicate a persistence of negative affect throughout the lab session for participants who received the painful interruption cue. These participants also experienced lower pleasure and higher arousal than participants who received the non-painful interruption cues, but only immediately after the calibration phase, pointing to a potential novelty effect of the painful stimulus that wore out during the experimental task.
Although interrupting ongoing activities is a natural response to pain , the effects of activity interruptions by pain on the subsequent pattern of activity performance have been scarcely investigated. The aim of the present experiment was to shed light on how people perform simple tasks that are interrupted by pain. Healthy volunteers performed an ongoing task that was occasionally interrupted by either a painful electrocutaneous stimulus or a non-painful and non-aversive auditory stimulus, followed by an interruption task. The latter had a maximum, but not a fixed duration. We expected that participants whose activity was interrupted by pain would exhibit an activity performance pattern comprising longer breaks from the interrupted (and pain-relevant) activity, and would therefore also need more time to complete the activity, as compared to participants interrupted by non-painful stimuli. Further, they would report lower motivation to perform the interrupted activity.
Results can be readily summarized. The two groups did not differ with regards to time spent away from the interrupted ongoing task and total time to complete the experimental task (i.e. the ongoing task including the time on the interruption task). Reported motivation for the interruption task was somewhat higher than for the ongoing task, but did not differ between the two groups. Our hypotheses were thus not confirmed. Further, in contrast to our expectations and to a previous study , pain catastrophizing did not emerge as an important factor in the consequences of interruptions by pain on performance. Its potential role thus warrants further investigation.
A theoretical model of activity interruptions by pain predicted that these will have negative consequences . Further, previous research has shown that people with pain respond to interruptions of work-related activities by continuing these activities outside work hours ; in other words, by adjusting the way in which they allocate time to their activities that are interrupted by pain. Our results, however, suggest that activity interruptions by pain do not necessarily lead to adjustments in the pattern of activity performance more than activity interruptions by pain-irrelevant stimuli do. Our finding also deviates from existing evidence that pain impairs concurrent task performance more than control stimuli (such as non-aversive somatic stimulation , or aversive auditory stimulation ).
On the other hand, our finding is in line with recently published studies showing that activity interruptions by pain and by non-pain impair quality of performance (expressed in accuracy and reaction times) to a similar degree , . One explanation might be that pain affects task performance only instantly, but does not have a delayed effect . Further, it is possible that people are to a certain degree resilient to the negative consequences of activity interruptions by pain. That might especially be the case in a group of young and healthy participants and when the interrupted task is fairly easy, as in the present study. Studies from outside the field of pain have shown that, even though interruptions are mostly disruptive , , , they may sometimes improve performance, especially when the interrupted task is very easy, boring or repetitive , , as the current ongoing task might have been. Although recent data shows that interruptions by pain are not more disruptive for the performance of a fairly complex activity compared to interruptions by non-pain , the same might not be true as regards performance pattern.
Another task characteristic might also explain our results, at least partly. Specifically, the completion point of our ongoing task was clearly defined, and could not be avoided unless the participant chose to terminate the lab session. This is in contrast to the open-ended ongoing task with one interruption of fixed duration that was used in the study of Schrooten and colleagues . It is plausible that when a task must be completed but is interrupted by pain, and the option to complete it by different means (e.g. by delegating it to another person) is unavailable, the person tries to complete it efficiently by avoiding long breaks that might impair performance and cause further delays .
The expected negative consequences of activity interruptions by pain might not necessarily be visible in the performance of the interrupted task, for example because people anticipate the negative effects and try to compensate for them . Research from outside the field of pain suggests that performance decrements caused by interruptions can be prevented and even reversed if one adapts their working strategy, but that this can lead to the experience of higher stress, frustration, annoyance and perceived effort invested in the task , , . In the same manner, interruptions by pain might not necessarily impair performance accuracy or performance pattern more than interruptions by non-pain, as this and previous research has shown , , but may have (more) negative emotional consequences. The inherent unpleasantness of pain  in combination with the (potential) use of compensatory strategies to prevent anticipated performance decrements might give rise to negative affect after interruptions by pain. In partial support of this idea, our exploratory data shows that negative affect persisted across the lab session in participants who were interrupted by pain, whereas it decreased in participants who were interrupted by the non-painful stimuli. Future research can further delve into this topic, for example by measuring specific affective states (e.g. specifically frustration rather than generically negative affect), by measuring affect immediately after an interruption (instead of upon completion of the whole activity), and/ or by disentangling the affective consequences of activity interruptions by pain from these of pain itself (e.g. by comparing activity interruptions by pain with pain that is not followed by an interruption).
The observation that participants preferred the interruption task over the ongoing task can be explained by task characteristics. The repeated interruptions that occurred during the ongoing task might have rendered it more unpleasant, and thus less motivating, than the interruption task, irrespective of the type of stimulus that prompted the interruption. Further, participants might have found the rapid change of screen configuration in the interruption task more interesting than the ongoing task, which might have been perceived as slower. The role of the motivational value of both the interrupted and the interrupting task is an interesting avenue for future research. Recent studies indicate that pain-related avoidance is reduced in the presence of valued goals , . As pain does not occur in a motivational vacuum, it would be informative to investigate task preferences (e.g. in terms of time spent on task) as a function of the motivational value of the interrupted and the interrupting task.
This study is not without limitations. Firstly, as mentioned above, the ongoing task was relatively easy. It is possible that people interrupted by pain take longer breaks when the task imposes higher (cognitive) load. Also, we considered physical tasks to be relevant to the daily life of people with pain and thus explicitly aimed at simulating a motoric task with a clear completion point. However, our paradigm is a laboratory task and might thus lack in ecological validity. Future research may investigate task type and/or load as factors moderating the effects of interruptions by pain . Secondly, the non-pain control stimulus that we used was a bell sound that differed from the painful stimulation not only in aversiveness, but also in modality. A control stimulus that differs with respect to only one dimension might have been preferred. Previous research, though, has provided evidence for the pain-specificity of the interference effects of pain. For example, Forkmann and colleagues  found that pain causes greater impairment in visual encoding than auditory stimuli matched to pain for unpleasantness. In any case, it is recommended for future research to use control stimuli matched to pain in as many dimensions possible. Similarly, future research may replicate the present study with tonic pain stimuli (such as heat pain), which may approach the chronic pain experience more closely. Third, our findings might not readily generalize to clinical samples, because our sample consisted of healthy volunteers. Further, as our study was adequately powered to detect large effects, it is possible that small or moderate effects exist, but could not be captured in this experiment. It would be worthwhile for future research to replicate the current findings with bigger and potentially more diverse samples.
To our knowledge, this is the first experiment that compared interruptions by painful stimuli to interruptions by non-painful stimuli with regards to their effect on the pattern of activity performance. Although activity interruptions by pain probably contribute to (chronic) pain patients’ disability to a significant degree, they have not been systematically investigated. The findings of research on activity interruptions by pain can potentially be used for the better management of interruptions by pain. This severely understudied area deserves scientific attention.
The authors wish to thank Jeroen Clarysse for technical support.
Boehm-Davis DA, Remington R. Reducing the disruptive effects of interruption: a cognitive framework for analysing the costs and benefits of intervention strategies. Accid Anal Prev 2009;41:1124–9. Web of SciencePubMedCrossrefGoogle Scholar
Adamczyk PD, Bailey BP. If not now when? The effects of interruption at different moments within task execution. In: proceedings of the SIGCHI conference on human factors in computing systems [Internet], August 1–4, Cambridge, Massachusetts. ACM 2004;271–8. Available at: http://portal.acm.org/citation.cfm?id=985692.985727.
Mark G, Gudith D, Klocke U. The cost of interrupted work: more speed and stress. In: proceedings of the SIGCHI Conference on Human Factors in Computing Systems, April 5–10, Florence, Italy. 2008;107–10. Google Scholar
Westbrook JI, Woods A, Rob MI, Dunsmuir WTM, Day RO. Association of interruptions with an increased risk and severity of medication administration errors. Arch Intern Med 2010;170:683–90. CrossrefPubMedWeb of ScienceGoogle Scholar
Gatzounis R, Schrooten MGS, Crombez G, Vancleef LMG, Vlaeyen JWS. Taking a break in response to pain. An experimental investigation of the effects of interruptions by pain on subsequent activity resumption. Scand J Pain 2017;16:52–60. Web of SciencePubMedGoogle Scholar
Gatzounis R, Schrooten MGS, Crombez G, Vlaeyen JWS. Activity interruptions by pain impair activity resumption, but not more than activity interruptions by other stimuli. An experimental investigation. Pain 2018;159:351–8. CrossrefGoogle Scholar
Meulders A, Franssen M, Fonteyne R, Vlaeyen JWS. Acquisition and extinction of operant pain-related avoidance behavior using a 3 degrees-of-freedom robotic arm. Pain 2016;157:1094–104. PubMedWeb of ScienceCrossrefGoogle Scholar
Hockey GRJ. Cognitive-energetical control mechanisms in the management of work demands and psychological health. In: Baddeley AD, Weiskrantz L, editors. Attention: selection, awareness, and control: a tribute to Donald Broadbent. New York, NY, US: Clarendon Press/Oxford University Press, 1993:329–45. Google Scholar
Meulders A, Vansteenwegen D, Vlaeyen JWS. The acquisition of fear of movement-related pain and associative learning: a novel pain-relevant human fear conditioning paradigm. Pain 2011;152:2460–69. Web of ScienceCrossrefPubMedGoogle Scholar
Van Damme S, Crombez G, Bijttebier P, Goubert L, Van Houdenhove B. A confirmatory factor analysis of the Pain Catastrophizing Scale. Invariant factor structure across clinical and non-clinical populations. Pain 2002;96:319–24. CrossrefPubMedGoogle Scholar
Engelen U, De Peuter S, Victoir A, Van Diest I, Van Den Bergh O. Verdere validering van de Positive and Negative Affect Schedule (PANAS) en vergelijking van twee Nederlandstalige versies. [Further validation of the Positive and Negative Affect Schedule (PANAS) and comparison of two Dutch versions]. Gedrag Gezond 2006;34:89–102. Google Scholar
Rhudy JL, France CR, Bartley EJ, Williams AE, McCabe KM, Russell JL. Does pain catastrophizing moderate the relationship between spinal nociceptive processes and pain sensitivity? J Pain 2009;10:860–9. CrossrefPubMedWeb of ScienceGoogle Scholar
van der Werf SP, Vos S. The prospective retrospective memory questionnaire: psychometric properties and normative data of a Dutch translation. Neth J Psychol 2011;66:120–8. Google Scholar
Karsdorp PA, Vlaeyen JWS. Goals matter: both achievement and pain-avoidance goals are associated with pain severity and disability in patients with low back and upper extremity pain. Pain 2011;152:1382–90. CrossrefPubMedWeb of ScienceGoogle Scholar
Spruyt A, Clarysse J, Vansteenwegen D, Baeyens F, Hermans D. Affect 4.0: a free software package for implementing psychological and psychophysiological experiments. Exp Psychol 2009;57:36–45. Web of ScienceGoogle Scholar
LimeSurvey Project Team, Schmitz C. LimeSurvey: An Open Source survey tool [Internet]. LimeSurvey Project Hamburg, Germany.; 2012. Available at: http://www.limesurvey.org.
Howell DC. Statistical methods for psychology, 6th ed. Belmont, CA: Thomson Wadsworth, 2007. Google Scholar
Field A. Discovering statistics using SPSS, 2nd ed. London: Sage publications, 2005. Google Scholar
IBM. IBM SPSS Statistics for Windows. Armonk: IBM Corporation, 2013. Google Scholar
Faul F, Erdfelder E, Lang A-GG, Buchner A, Kiel C. G*Power3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175–91. PubMedCrossrefGoogle Scholar
Forkmann K, Wiech K, Ritter C, Sommer T, Rose M, Bingel U. Pain-specific modulation of hippocampal activity and functional connectivity during visual encoding. J Neurosci 2013;33:2571–81. CrossrefPubMedWeb of ScienceGoogle Scholar
Ariga A, Lleras A. Brief and rare mental “breaks” keep you focused: deactivation and reactivation of task goals preempt vigilance decrements. Cognition 2011;118:439–43. PubMedWeb of ScienceCrossrefGoogle Scholar
Merskey H, Bogduk N. Classification of chronic pain [Internet]. Second. In: Merskey H, Bogduk N, editors. Classification of chronic pain. Seattle: IASP Press, 1994. Available at: http://www.iasp-pain.org/Taxonomy#Pain.
The online version of this article offers supplementary material (https://doi.org/10.1515/sjpain-2017-0098).
These comprised a Situational version of the Pain Catastrophizing Scale (Pain participants only; ) within the session, and the Fear of Pain Questionnaire , , Prospective and Retrospective Memory Questionnaire , , Goal Pursuit Questionnaire , Prevention/Promotion Scale  and Positive and Negative Affect Schedule, Trait version, in an online questionnaire session that took place 1 week after the participant came to the lab.
About the article
Published Online: 2018-02-13
Published in Print: 2018-01-26
Research funding: This study was supported by a PhD “Aspirant” grant (grant ID 11N8215N) granted to RG by the Research Foundation – Flanders, Belgium (FWO Vlaanderen). JWSV was supported by the Odysseus Grant “The Psychology of Pain and Disability Research Program” (grant ID G090208N) funded by the Research Foundation Flanders (FWO Vlaanderen), Belgium, and the “Asthenes” (# METH/15/011) long-term structural funding-Methusalem grant by the Flemish Government, Belgium.
Conflict of interest: All authors have discussed the results and commented on the manuscript. The authors have no conflict of interest to report. We report all data exclusions, all manipulations, and all measures in the study.
Informed consent: Participants provided informed consent prior to participation.
Ethical approval: The study described in the present manuscript was approved by the appropriate Ethical Board (Social and Societal Ethics Committee, reg. nr. S54157, and Medical Ethics Committee of the University of Leuven, reg. nr. ML8125).
Citation Information: Scandinavian Journal of Pain, Volume 18, Issue 1, Pages 109–119, ISSN (Online) 1877-8879, ISSN (Print) 1877-8860, DOI: https://doi.org/10.1515/sjpain-2017-0098.