Behav Res DOI 10.3758/s13428-014-0453-3
Measuring mechanical pain: The refinement and standardization of pressure pain threshold measurements Michael Melia & Martin Schmidt & Britta Geissler & Jochem König & Ulrike Krahn & Hans Jürgen Ottersbach & Stephan Letzel & Axel Muttray
# Psychonomic Society, Inc. 2014
Abstract Pain thresholds are widely used in behavioral research, but unlike other pain modalities, a standardized assessment of pressure pain remains a challenge. In this research, we describe the application of an automatic pressure algometer with a linear increase in force. Ergonomically designed fixation devices were developed to increase the accuracy and to shorten the time of each measurement. Ten healthy volunteers were included in a pilot study to test the algometry method. Pressure pain thresholds (PPTs) were investigated over 2 experimental days in three nonconsecutive runs at 29 measurement sites. During the experiment, subjects reported their subjective sleepiness, level of state-anxiety, psychological status and the perceived pain intensity of each measurement. Pain intensity ratings indicate that instructions were followed. State-anxiety and subjective sleepiness levels were low throughout the experiment. The method has proven to be suitable for standardized PPT measurements across the body in an ergonomic, safe, and user-friendly fashion.
Keywords Algometer . Pressure pain threshold . Standardization
Contains data from the med. dissertation in preparation of M. Schmidt. M. Melia (*) : M. Schmidt : B. Geissler : S. Letzel : A. Muttray Institute of Occupational, Social and Environmental Medicine, 55131 Mainz, Germany e-mail: [email protected] J. König : U. Krahn Institute of Medical Biostatistics, Epidemiology and Informatics, Division Medical Biometry, University Medical Center Johannes Gutenberg University, 55131 Mainz, Germany H. J. Ottersbach Institute for Occupational Safety of the German Social Accident Insurance (IFA), 53757 Sankt Augustin, Germany
Introduction Today, there are many instruments, devices, and methods that can be used to evaluate the perception of pain. The range of testable modalities includes thermal, electrical, chemical, and mechanical stimuli, of which the measurement of mechanical pain is the most widely used in research (Arendt-Nielsen & Yarnitsky, 2009). In recent years, the demand for standardized computer-controlled pressure stimulation in behavioral research has risen. Pressure algometry methods can be used for clinical research (Dannecker & Sluka, 2011; Egloff et al., 2011; Norregaard, Jacobsen, & Krisrensen, 1999; Rolke et al., 2006), for measuring the efficacy of therapeutic interventions for the treatment of pain (Moraes Maia et al., 2012; Olesen et al., 2013; Skou et al., 2012), and for general psychophysiological research (Bohns & Wiltermuth, 2012; Dunbar et al., 2012; Pollatos, Füstös, & Critchley, 2012). Apart from some perceptional influencing factors, the psychophysiological assessment of pressure pain thresholds (PPTs) highly depends on the kind of device used (Rolke et al., 2006), on the increase in force (Dyck et al., 1990; Dyck et al., 1993; List, Helkimo, & Karlsson, 1991; Shy et al., 2003), and on the characteristics of the respective contact surface (Defrin, Ronat, Ravid, & Peretz, 2003; Greenspan & McGillis, 1991). Available pressure algometers range from manually operated spring-loaded systems (Jensen, Andersen, Olesen, & Lindblom, 1986) to automatic systems, which, in turn, can be subdivided into constant-velocity-controlled devices (Saito & Ikeda, 2005) and force-controlled algometers (Adnadjevic & GravenNielsen, 2012; Zimkowski, Lindley, Patel, & Rentschler, 2011). For reasons of simplicity, the most commonly used devices are hand-held algometers with a 1-cm2 pressure application surface (Kinser, Sands, & Stone, 2009). However, one of the biggest drawbacks of hand-held algometers is that it is difficult to maintain consistent application rates, especially over
Behav Res
multiple test periods. Reported pressure application rates range from 0.05 to 20 N/s, whereby faster rates can provoke a false low-threshold reading (Jensen, Andersen, Olesen, & Lindblom, 1986). The main problem surrounding automatic distance-controlled systems is that pressure exertions around high-resisting tissue (e.g., bone) can potentially create harmful pressure peaks. Some studies report unusually fast application rates causing potential response time errors, experimental anxiety, and at worst, tissue damage (Hägg & Astrom, 1997; Kadetoff & Kosek, 2007; Kosek & Lundberg, 2003). Furthermore, different tissue properties may also create nonlinear increases in force and, ultimately, inconsistent presentations of stimuli. Due to the high intraindividual variability in the perception of pressure stimuli, it is important to create reproducible conditions, especially for multiple testing phases at different sites and at different times. Another problem lies in the insufficient recording of data. Many algometers display only the maximum force reached during a measurement. Therefore, the course of pressure is rarely investigated (Kinser et al., 2009). However, force time functions are of great importance for reliable conclusions on the linearity of pressure application rates. All in all, it was our objective to account for these potential confounders and to evaluate possible health risks when measuring PPTs at 29 different measurement sites, using a fully automated algometer with a quasi-linear increase in force. Pain threshold is defined as the minimum intensity of a stimulus (e.g., pressure) that is perceived as painful and functions as a warning system in a self-protecting and adapting organism (Gibson & Farrell, 2004). However, unlike for other modalities (e.g., heat), standardizing pain thresholds for pressure has proven to be more challenging. In the following section, the details of a refined method of standardizing PPTs will be reported with some exemplary results from a pilot study. The aim was to test complete runs of measurements, including the feedback of test subjects, and to obtain a preliminary overview of individual differences in PPTs across the body. Due to the underlying questions of this project and some ethical implications, we decided not to measure pressure pain tolerance.
Method Algometer The test facility, which was technologically developed by the German Institute for Occupational Safety and Health (IFA) (Fig. 1a–d), includes a fully automatic pressure algometer, a PC-station (Intel ® Core i5 CPU 3.46 GHz; Windows 7), as well as an apparatus for positioning selected body sites with different adjustable devices. Running the algometer requires two programs, both of which were also developed by the IFA.
a
b
c
d
Fig. 1 Test facility “pressure algometer.“ a Measurement of the chest bone (sternum). b Measurement of the pelvis (Spina iliaca ndom). c Measurement of the fingertip (Dip II ndom). d Measurement of the spinous process C 7 (Proc. spinosus vertebrae cervicalis 7). The response key is held in in the right or left hand. The test subject is blindfolded. The algometer was built by the Institute for Occupational Safety and Health of the German Social Accident Insurance (IFA)
The control system software is responsible for achieving a linear increase in force, and the investigator program functions as a database for all measurements. Relevant boundary parameters and measurement files are stored in that database. The algometer is fixated in a gantry structure that can be shifted vertically across the whole body length and is placed on a three-axis mechanism delivering large kinematic degrees of freedom. Automatic digital displays provide the correct geometric position, creating the highest possible level of spatial reproducibility. The plunger can be chosen and adapted according to the specific purpose of the experiment. Plungers of different sizes, shapes, or materials can be clamped in a threaded hole and rotated 360°. Once the measurement is started, the control program issues the command for the plunger to advance forward toward the respective measurement site. This is accompanied by a beeping sound. The plunger moves forward with a constant velocity of 0.5 mm/s until 5 N of force is achieved. Subsequently, it moves back with the same velocity until a contact force of 2 N is undershot for the first time, initiating
Behav Res
the measurement, which is signaled by a second beeping sound. The purpose of this first contact phase is to set the distance control system to zero, for an accurate distance measurement, and to synchronically initiate the adapted force control system. During the load application, the algometer adjusts the approximated linear increase in force applied by taking the progression of distance into account. Small discrete steps come about as a result of stepwise adjustments to the respective tissue properties. The linear motor ideally readjusts itself every 10 ms. Due to the deformation behavior of the elastic properties of the respective material and the inertia behavior of the inner masses of the linear motor during the control process, a time period of approximately 50 ms is reached for the adjustment of the linear force increase.The pressure algometer can generate a maximum applied force of 500 N, and force increases from 1 to 10 N/s. The test subject is able to terminate the application of force at any chosen time by activating a response key (Fig. 1a–c), which will lead to a rapid returning of the algometer to its home position. This enabling device has three independent positions: (1) open/released position, (2) intermediate position, and (3) fully depressed position. The investigator is also provided with a response key of the same type and with the same function. Pressure exertion is initiated and resumed if both enabling devices are held in the intermediate position and continues until one of the response keys is released or fully depressed. The total force distributed over the surface of the plunger is recorded over time by the algometer with a sampling rate of 100 Hz. The corresponding parameters registered by a PC station include time (ms), distance covered (mm), maximum force recorded (N), and the force during response key activation. Tables and graphs are automatically stored for later evaluation. The software also allows for boundary parameters (Fig. 2), such as personal information of the test subject, exact localization of the measurement site, temperature, and humidity, as well as test observations made by the investigator, to be saved. The algometer is equipped with a redundantly performing electromechanical safety device. Two independent systems reliably monitor the overall operation of the algometer, ensuring that an upper force limit is not exceeded. This force limit is set up at first manually with a mechanical spring-loaded system. The preloaded force is metrologically monitored by a second integrated electronic system, which is dually channeled. This integrated electronic system has the function of guaranteeing the safe monitoring of the upper force limit on the basis of the precisely well-known force displacement function of the spring system. The upper force limit chosen for this pilot study was 150 N for most measurement points and 50 N for three sensitive points (i.e., cranial sites). Once the upper force limit is reached, the algometer immediately falls back and assumes a safety position. In general, once a measurement is terminated by the test subject, the safety control system automatically assumes that there is a reduction
in force. If, however, a countermovement is made by the subject after pressing the response key, the inconsistency will be registered, and the algometer will also fall back into its safety position. This is mostly the case for sites that are prone to movement. Furthermore, the algometer can be turned off at any time, using one of the three emergency halt buttons. A Declaration of Conformity was drawn up for the entire test facility, in compliance with the technical safety requirements of the German Machinery Directive, and a CE mark was assigned. Pretests Months prior to the beginning of the pilot study, a series of preliminary tests were conducted—including test measurements with 13 volunteers—for the refinement of the algometer and the development of a standardized procedure. Different plunger sizes and shapes were tested until the final one was selected. The plunger used for the pilot study was plastic and square in shape. To prevent the perception of unwanted sharp edges, the sides were rounded off with a radius of 2 mm. The flat surface measured 10×10 mm. The specific shape was chosen for the adaptation of a specially designed pressureindicating film. The film is still in a developmental stage and is expected to be implemented in future studies. Selection of measurement sites In total, 29 measurement sites (Fig. 4; 28 localizations with females, excluding the breast muscle) were selected. Each measurement site was allocated to one of five body regions: “head and neck,” “trunk,” “arm,” “upper extremities,” and “lower extremities.” All sites were marked according to an exact medical definition of the localization. With regard to prevention, most measurements were conducted on the nondominant (ndom) side, since some research suggests that lower pain thresholds might be attributed to the nondominant hemisphere of the body (Brennum, Kjeldsen, Jensen, & Staehlin Jensen, 1989; Buchanan & Midgley, 1987; Göbel & Westphal, 1987; Özcan, Tulum, Pinar, & Başkurt, 2004; Spernal, Krieg, & Lautenbacher, 2003). Four sites were selected on the dominant side of the hand in order to draw comparisons with their nondominant counterparts. Development of fixation devices Considering the large number of 29 measurement sites, which are spread across the body, an ergonomic and time efficient method had to be created. Furthermore, finding the same localization is crucial for PPT measurements, due to the large variability and necessity of repeated measurements. To achieve reproducible positions, we used an array of different fixation devices (Fig. 1), including vacuum cushions (Schmidt GmbH, Garbsen, Germany), which assume the shape of the
Behav Res
Fig. 2 Input mask for boundary conditions
respective body part, adjustable hook and loop fasteners, chinrests, and metal panels in different sizes and shapes (Item GmbH, Solingen, Germany). The primary goal was to reach any part of the body, fixated in a comfortable but motion-resistant position. Test subjects were positioned while seated or standing with their individualized vacuum cushions placed on a metal panel in front of a large-surface and rotatable wall, embedded in a framework structure. Folder and bar profiles were set at predefined heights and widths (depending on body size and shape), with the vacuum cushions keeping the desired body part in the required position (Fig. 1). All settings were then recorded. After the measurements test, subjects were asked to give their feedback on the fixation devices.
Experimental study design In a pilot study, PPTs of 10 volunteers were measured across the body. Localizations were combined to five measurement blocks
Fig. 3 Timeline of experimental procedure
(consisting of 1–10 measurement sites), which were randomized. The order of this block-wise randomization was kept for each cycle. Each localization was measured three times, nonconsecutively. All measurements were spread over 2 days.
Subects Ten healthy volunteers (6 male, 4 female; median age = 29 years; range = 23–67 years), predominantly students, were included in the study after proving eligibility in an anamnestic and physical internal and orienting neurological examination. We screened for illnesses that could present a risk for test subjects or alter the results of the measurements. These include psychiatric, neurological, and chronic pain illnesses, as well as movement and coagulation disorders. A tendency to bleed was excluded with a validated questionnaire (Koscielny et al., 2004). None of the subjects had a recent history of musculoskeletal pain or took pain medication. One female was on an oral contraceptive. All 4 females were not having
Behav Res
Head and Neck
Trunk
Arm
Hand
Leg
Fig. 4 Overview of all 29 measurement sites. ndom = non-dominant side; dom = dominant side
their period. A male subject took finasterid because of a prostate hypertrophy but did not report any side effects. All volunteers were right-handed pursuant to the Edinburgh Handedness Inventory (Oldfield 1971). None of the subjects were obese (mean body mass index [BMI], 22.9 kg/m2; range, 19.4–27.5 kg/m2). One subject was a smoker. The study was approved by the local ethics committee and was conducted in accordance with the declaration of Helsinki and its amendments. Informed written consent was obtained from all volunteers prior to the inclusion of the study. All subjects received a monetary compensation for their participation. Pain perception and self-assessment of subjects After each measurement test, subjects were asked to place a mark on a 100-mm visual analog scale (VAS) to indicate the subjective perception of pain intensity at the measurement site ranging from 0 mm = no pain to 100 mm = most pain imaginable. The State Anxiety Inventory (STAI-S; Spielberger, 1983) was used to measure possible experimental anxiety. The
inventory is easy to score and possesses good validity and internal reliability (Cronbach’s α=.91) (Spielberger, 1983). The Brief Symptom Inventory (BSI) is a 53-item selfreport questionnaire designed to reflect the current psychological symptom status (Derogatis, 1975). Each item is rated on a 5-point Likert-type scale ranging from not at all to extremely. The items define a broad spectrum of perceived restrictions relative to physical and psychological symptoms that occurred in the preceding 7-day period. Symptoms are assigned to nine subscales, which represent domains of psychopathology (e.g., trait-anxiety, depression, and somatization). The Karolinska Sleepiness Scale (KSS) was used for evaluating possible subjective sleepiness, using a 9-point scale, ranging from extremely alert to extremely sleepy/fighting sleep (Åkerstedt & Gillberg, 1990). Experimental procedure The first day began with a medical examination to check for possible pain-related disorders and current illnesses (Fig. 3).
Behav Res
Afterward, subjects filled out the BSI and were introduced to the experiment. Initially, three test measurements were conducted using a conventional hand-held pressure algometer (Sauter FA 500, Greenwich, U.S.), followed by seven test measurements with the automatic algometer (Fig. 1). The primary goal was to approach the concept of pain thresholds and to reduce discomfort and possible anxiety. The STAI-S was filled out four times: (1) during the medical examination, (2) after the algometer room was entered for the first time, (3) before the first experimental measurement, and (4) after the last experimental measurement. Under certain circumstances, the investigator’s gender can influence the reaction of the subject during the experiment (Levine & De Simone, 1991). This is why the investigator was always of the same sex as the subject. Interpersonal expectancy effects (Rosenthal & Rubin, 1978) were minimized by means of neutral and standardized instructions and by practicing the algometric procedure. The measurements took place the next day and were spread over 2 experimental days. The average temperature was 23.1 °C (SD = 0.81), and the relative air humidity was 36.9 % (SD=6.6). Subjects were blindfolded for each measurement. Pressure was applied perpendicular to the skin. The PPT was signaled by pressing or releasing the response key and was defined by the following instruction: “Please press or let go, when pressure transitions into beginning pain.” The increase of pressure was set at 5 N/s, except for the chewing muscle (M. masseter ndom) and temple (Os temporale ndom), which were set at 2 N/s owing to low PPTs at those sites, which were obtained during our preliminary experiments. After each measurement, subjects immediately recorded their perceived level of pain using the VAS. Subjects were also asked to report their personal observations (e.g., unusual sensations, lingering pain, etc.). No lingering pain was reported. If the pressure was not distributed evenly on the skin, the subject was asked to identify the exact point where possible edges or other special features were felt. Those measurements were repeated at the end of all regular measurements. Breaks were taken on the basis of the subject’s individual well-being and his or her vigilance. Statistical methods Due to the small sample size of this pilot study, data are evaluated exploratively and are presented descriptively. According to an analysis of distribution, PPTs (and pain intensity ratings) are log-normal distributed. As a measure of central tendency, we used the mean value of three measurements for single localizations of every subject (Hooten, Rosenberg, Eldrige, & Qu, 2013; Zolnoun et al., 2012). The geometric mean was used for aggregating the values of all subjects and/or measurement sites. In order to quantify how the variation of the PPT, as well as the rating of pain intensity (VAS), was distributed over the
factors subject, measurement localization, and random error, we fitted a hierarchical linear random effects model. Spearman’s rank correlation coefficient with the corresponding 95 % confidence interval (95 %-CI) was used for the analysis of bivariate correlations, and the Friedman test was used to compare repeated measurements. All presented p values are two-sided. Boxplots and an illustration of normalized geometrical means with corresponding confidence intervals were used for the graphical presentation. The analysis was performed using the statistical software SAS version 9.3 (SAS, Cary, NC), SPSS version 20 (IBM, Armonk, NY), and R version 2.12.2 (R Development Core Team).
Results Functioning and safety of test facility The complete test facility has proven to be reliable, easy to operate, and durable in all phases of the design and testing process. During the pilot study, no major technical disturbances occurred in the operation of the automatic pressure algometer. The increase in pressure was, for almost all measurements, quasi-linear (Fig. 5). Single fluctuations of force were observed as a result of different tissue properties or movement caused by breathing. Including our preliminary tests, we conducted roughly 1,500 PPT measurements and found no damaging effects on the skin. Even after repeated measurements at various sites, neither skin irritations nor local hematomas were observed. During our preliminary tests, we discovered that measuring 29 sites is very time consuming. Despite using adjustable fixation devices, one cycle lasted up to 5.5 h. The more efficient approach as described for the pilot study of grouping measurement sites into blocks of neighboring localizations reduced each cycle, on average, by 2 h. The mean time of 1 experimental day was 5 h and 19 min (SD=24 min). By minimizing the amount of reconstructions of our fixation devices, we saved 6 h per subject, on average, for the complete experiment. The vacuum cushions adapted perfectly to all anatomical structures. All fixation devices were ergonomic and comfortable, according to the direct feedback of our volunteers. One subject described the measurements at the kneecap and another at the temple as unpleasant. Four out of 10 subjects reported some difficulties in defining the onset of pain during the test measurements but were able to make correct assessments during the experiment. Pressure pain thresholds The average PPTs at 29 measurement sites varied among the subjects (Fig. 6). With the exception of the forehead, low PPTs
Behav Res
were predominantly recorded at or near cranial sites, which include the chewing muscle, the temple, and the neck muscle. The highest PPTs were measured at the palm of the hand, the index finger tip, and the calf muscle.
[N]
100
Pain perception and self-assessment of subjects
80
60
40
20
0 0
5
10
15
20
25
[s]
Fig. 5 Force-time curve of a measurement at the calf muscle (ndom)
Fig. 6 Pressure pain thresholds (PPTs) of all localizations. PPTs of 10 test subjects were averaged over three repeated measurements. Boxplot: The 1st quartile value is at the lower horizontal line of the box; the 3rd quartile value is at the upper horizontal line; the thick line in the box represents the median. The whiskers (the lines that extend out the top and bottom of the box) represent the highest and lowest values that are not outliers. Outliers (values that are between 1.5 and 3 times the interquartile range) are represented by circles beyond the whiskers. The breast muscle was measured only with male subjects (n=6)
The median pain intensity score after measuring the PPT was 6 mm (min = 0 mm, 2nd quartile = 3 mm, 3rd quartile = 11 mm, max = 32 mm). The maximum rating of 32 mm was recorded after a measurement of the palm of the hand on the nondominant side (54.9 N). Figure 7 shows a tendency for higher pain intensity ratings at more sensitive measurement sites (e.g., chewing muscle) and vice versa (p=.051, effect of VAS in a random slope mixed model with VAS as the only fixed effect). This was no longer the case after accounting for the measurement sites (p=.26). Within all 29 measurement sites, only the pain intensity rating of the palm of the hand (ndom) showed an inverse relationship with the corresponding PPT (Spearman’s rho = −.37, 95 %-CI [−0.64; −0.24], p=.046). On the basis of the observations of all experimental days, the median score of the KSS was 3 (min = 2, 1st quartile = 3, 3rd quartile = 4, max = 6), indicating an acceptable level of vigilance. Scores of 6 (between neither awake nor tired and tired but no problems staying awake) were reported twice. Scores decreased in both cases after having taken a break. State-anxiety scores of our subjects were generally low. The overall level of state-anxiety changed only marginally on a group basis during the four different time points, χ 2 (3) = 1.16, p = .76 (Friedman test) (Fig. 8). Two subjects revealed slightly higher scores, one of which exceeded the postulated level of 50 during the medical examination, indicating an elevated stateanxiety level (Hermes, Matthes, & Saka, 2007), and the other after entering the algometry room for the first time. In both cases, state-anxiety normalized on the second day before the first round of measurements took place. The scores of the trait-anxiety subscale of the BSI correlated negatively with the aggregated geometrical mean PPT score of all measurement sites (Spearman’s rho = −0.71, 95 %CI [−0.94;−0.23], p
Measuring mechanical pain: The refinement and standardization of pressure pain threshold measurements Michael Melia & Martin Schmidt & Britta Geissler & Jochem König & Ulrike Krahn & Hans Jürgen Ottersbach & Stephan Letzel & Axel Muttray
# Psychonomic Society, Inc. 2014
Abstract Pain thresholds are widely used in behavioral research, but unlike other pain modalities, a standardized assessment of pressure pain remains a challenge. In this research, we describe the application of an automatic pressure algometer with a linear increase in force. Ergonomically designed fixation devices were developed to increase the accuracy and to shorten the time of each measurement. Ten healthy volunteers were included in a pilot study to test the algometry method. Pressure pain thresholds (PPTs) were investigated over 2 experimental days in three nonconsecutive runs at 29 measurement sites. During the experiment, subjects reported their subjective sleepiness, level of state-anxiety, psychological status and the perceived pain intensity of each measurement. Pain intensity ratings indicate that instructions were followed. State-anxiety and subjective sleepiness levels were low throughout the experiment. The method has proven to be suitable for standardized PPT measurements across the body in an ergonomic, safe, and user-friendly fashion.
Keywords Algometer . Pressure pain threshold . Standardization
Contains data from the med. dissertation in preparation of M. Schmidt. M. Melia (*) : M. Schmidt : B. Geissler : S. Letzel : A. Muttray Institute of Occupational, Social and Environmental Medicine, 55131 Mainz, Germany e-mail: [email protected] J. König : U. Krahn Institute of Medical Biostatistics, Epidemiology and Informatics, Division Medical Biometry, University Medical Center Johannes Gutenberg University, 55131 Mainz, Germany H. J. Ottersbach Institute for Occupational Safety of the German Social Accident Insurance (IFA), 53757 Sankt Augustin, Germany
Introduction Today, there are many instruments, devices, and methods that can be used to evaluate the perception of pain. The range of testable modalities includes thermal, electrical, chemical, and mechanical stimuli, of which the measurement of mechanical pain is the most widely used in research (Arendt-Nielsen & Yarnitsky, 2009). In recent years, the demand for standardized computer-controlled pressure stimulation in behavioral research has risen. Pressure algometry methods can be used for clinical research (Dannecker & Sluka, 2011; Egloff et al., 2011; Norregaard, Jacobsen, & Krisrensen, 1999; Rolke et al., 2006), for measuring the efficacy of therapeutic interventions for the treatment of pain (Moraes Maia et al., 2012; Olesen et al., 2013; Skou et al., 2012), and for general psychophysiological research (Bohns & Wiltermuth, 2012; Dunbar et al., 2012; Pollatos, Füstös, & Critchley, 2012). Apart from some perceptional influencing factors, the psychophysiological assessment of pressure pain thresholds (PPTs) highly depends on the kind of device used (Rolke et al., 2006), on the increase in force (Dyck et al., 1990; Dyck et al., 1993; List, Helkimo, & Karlsson, 1991; Shy et al., 2003), and on the characteristics of the respective contact surface (Defrin, Ronat, Ravid, & Peretz, 2003; Greenspan & McGillis, 1991). Available pressure algometers range from manually operated spring-loaded systems (Jensen, Andersen, Olesen, & Lindblom, 1986) to automatic systems, which, in turn, can be subdivided into constant-velocity-controlled devices (Saito & Ikeda, 2005) and force-controlled algometers (Adnadjevic & GravenNielsen, 2012; Zimkowski, Lindley, Patel, & Rentschler, 2011). For reasons of simplicity, the most commonly used devices are hand-held algometers with a 1-cm2 pressure application surface (Kinser, Sands, & Stone, 2009). However, one of the biggest drawbacks of hand-held algometers is that it is difficult to maintain consistent application rates, especially over
Behav Res
multiple test periods. Reported pressure application rates range from 0.05 to 20 N/s, whereby faster rates can provoke a false low-threshold reading (Jensen, Andersen, Olesen, & Lindblom, 1986). The main problem surrounding automatic distance-controlled systems is that pressure exertions around high-resisting tissue (e.g., bone) can potentially create harmful pressure peaks. Some studies report unusually fast application rates causing potential response time errors, experimental anxiety, and at worst, tissue damage (Hägg & Astrom, 1997; Kadetoff & Kosek, 2007; Kosek & Lundberg, 2003). Furthermore, different tissue properties may also create nonlinear increases in force and, ultimately, inconsistent presentations of stimuli. Due to the high intraindividual variability in the perception of pressure stimuli, it is important to create reproducible conditions, especially for multiple testing phases at different sites and at different times. Another problem lies in the insufficient recording of data. Many algometers display only the maximum force reached during a measurement. Therefore, the course of pressure is rarely investigated (Kinser et al., 2009). However, force time functions are of great importance for reliable conclusions on the linearity of pressure application rates. All in all, it was our objective to account for these potential confounders and to evaluate possible health risks when measuring PPTs at 29 different measurement sites, using a fully automated algometer with a quasi-linear increase in force. Pain threshold is defined as the minimum intensity of a stimulus (e.g., pressure) that is perceived as painful and functions as a warning system in a self-protecting and adapting organism (Gibson & Farrell, 2004). However, unlike for other modalities (e.g., heat), standardizing pain thresholds for pressure has proven to be more challenging. In the following section, the details of a refined method of standardizing PPTs will be reported with some exemplary results from a pilot study. The aim was to test complete runs of measurements, including the feedback of test subjects, and to obtain a preliminary overview of individual differences in PPTs across the body. Due to the underlying questions of this project and some ethical implications, we decided not to measure pressure pain tolerance.
Method Algometer The test facility, which was technologically developed by the German Institute for Occupational Safety and Health (IFA) (Fig. 1a–d), includes a fully automatic pressure algometer, a PC-station (Intel ® Core i5 CPU 3.46 GHz; Windows 7), as well as an apparatus for positioning selected body sites with different adjustable devices. Running the algometer requires two programs, both of which were also developed by the IFA.
a
b
c
d
Fig. 1 Test facility “pressure algometer.“ a Measurement of the chest bone (sternum). b Measurement of the pelvis (Spina iliaca ndom). c Measurement of the fingertip (Dip II ndom). d Measurement of the spinous process C 7 (Proc. spinosus vertebrae cervicalis 7). The response key is held in in the right or left hand. The test subject is blindfolded. The algometer was built by the Institute for Occupational Safety and Health of the German Social Accident Insurance (IFA)
The control system software is responsible for achieving a linear increase in force, and the investigator program functions as a database for all measurements. Relevant boundary parameters and measurement files are stored in that database. The algometer is fixated in a gantry structure that can be shifted vertically across the whole body length and is placed on a three-axis mechanism delivering large kinematic degrees of freedom. Automatic digital displays provide the correct geometric position, creating the highest possible level of spatial reproducibility. The plunger can be chosen and adapted according to the specific purpose of the experiment. Plungers of different sizes, shapes, or materials can be clamped in a threaded hole and rotated 360°. Once the measurement is started, the control program issues the command for the plunger to advance forward toward the respective measurement site. This is accompanied by a beeping sound. The plunger moves forward with a constant velocity of 0.5 mm/s until 5 N of force is achieved. Subsequently, it moves back with the same velocity until a contact force of 2 N is undershot for the first time, initiating
Behav Res
the measurement, which is signaled by a second beeping sound. The purpose of this first contact phase is to set the distance control system to zero, for an accurate distance measurement, and to synchronically initiate the adapted force control system. During the load application, the algometer adjusts the approximated linear increase in force applied by taking the progression of distance into account. Small discrete steps come about as a result of stepwise adjustments to the respective tissue properties. The linear motor ideally readjusts itself every 10 ms. Due to the deformation behavior of the elastic properties of the respective material and the inertia behavior of the inner masses of the linear motor during the control process, a time period of approximately 50 ms is reached for the adjustment of the linear force increase.The pressure algometer can generate a maximum applied force of 500 N, and force increases from 1 to 10 N/s. The test subject is able to terminate the application of force at any chosen time by activating a response key (Fig. 1a–c), which will lead to a rapid returning of the algometer to its home position. This enabling device has three independent positions: (1) open/released position, (2) intermediate position, and (3) fully depressed position. The investigator is also provided with a response key of the same type and with the same function. Pressure exertion is initiated and resumed if both enabling devices are held in the intermediate position and continues until one of the response keys is released or fully depressed. The total force distributed over the surface of the plunger is recorded over time by the algometer with a sampling rate of 100 Hz. The corresponding parameters registered by a PC station include time (ms), distance covered (mm), maximum force recorded (N), and the force during response key activation. Tables and graphs are automatically stored for later evaluation. The software also allows for boundary parameters (Fig. 2), such as personal information of the test subject, exact localization of the measurement site, temperature, and humidity, as well as test observations made by the investigator, to be saved. The algometer is equipped with a redundantly performing electromechanical safety device. Two independent systems reliably monitor the overall operation of the algometer, ensuring that an upper force limit is not exceeded. This force limit is set up at first manually with a mechanical spring-loaded system. The preloaded force is metrologically monitored by a second integrated electronic system, which is dually channeled. This integrated electronic system has the function of guaranteeing the safe monitoring of the upper force limit on the basis of the precisely well-known force displacement function of the spring system. The upper force limit chosen for this pilot study was 150 N for most measurement points and 50 N for three sensitive points (i.e., cranial sites). Once the upper force limit is reached, the algometer immediately falls back and assumes a safety position. In general, once a measurement is terminated by the test subject, the safety control system automatically assumes that there is a reduction
in force. If, however, a countermovement is made by the subject after pressing the response key, the inconsistency will be registered, and the algometer will also fall back into its safety position. This is mostly the case for sites that are prone to movement. Furthermore, the algometer can be turned off at any time, using one of the three emergency halt buttons. A Declaration of Conformity was drawn up for the entire test facility, in compliance with the technical safety requirements of the German Machinery Directive, and a CE mark was assigned. Pretests Months prior to the beginning of the pilot study, a series of preliminary tests were conducted—including test measurements with 13 volunteers—for the refinement of the algometer and the development of a standardized procedure. Different plunger sizes and shapes were tested until the final one was selected. The plunger used for the pilot study was plastic and square in shape. To prevent the perception of unwanted sharp edges, the sides were rounded off with a radius of 2 mm. The flat surface measured 10×10 mm. The specific shape was chosen for the adaptation of a specially designed pressureindicating film. The film is still in a developmental stage and is expected to be implemented in future studies. Selection of measurement sites In total, 29 measurement sites (Fig. 4; 28 localizations with females, excluding the breast muscle) were selected. Each measurement site was allocated to one of five body regions: “head and neck,” “trunk,” “arm,” “upper extremities,” and “lower extremities.” All sites were marked according to an exact medical definition of the localization. With regard to prevention, most measurements were conducted on the nondominant (ndom) side, since some research suggests that lower pain thresholds might be attributed to the nondominant hemisphere of the body (Brennum, Kjeldsen, Jensen, & Staehlin Jensen, 1989; Buchanan & Midgley, 1987; Göbel & Westphal, 1987; Özcan, Tulum, Pinar, & Başkurt, 2004; Spernal, Krieg, & Lautenbacher, 2003). Four sites were selected on the dominant side of the hand in order to draw comparisons with their nondominant counterparts. Development of fixation devices Considering the large number of 29 measurement sites, which are spread across the body, an ergonomic and time efficient method had to be created. Furthermore, finding the same localization is crucial for PPT measurements, due to the large variability and necessity of repeated measurements. To achieve reproducible positions, we used an array of different fixation devices (Fig. 1), including vacuum cushions (Schmidt GmbH, Garbsen, Germany), which assume the shape of the
Behav Res
Fig. 2 Input mask for boundary conditions
respective body part, adjustable hook and loop fasteners, chinrests, and metal panels in different sizes and shapes (Item GmbH, Solingen, Germany). The primary goal was to reach any part of the body, fixated in a comfortable but motion-resistant position. Test subjects were positioned while seated or standing with their individualized vacuum cushions placed on a metal panel in front of a large-surface and rotatable wall, embedded in a framework structure. Folder and bar profiles were set at predefined heights and widths (depending on body size and shape), with the vacuum cushions keeping the desired body part in the required position (Fig. 1). All settings were then recorded. After the measurements test, subjects were asked to give their feedback on the fixation devices.
Experimental study design In a pilot study, PPTs of 10 volunteers were measured across the body. Localizations were combined to five measurement blocks
Fig. 3 Timeline of experimental procedure
(consisting of 1–10 measurement sites), which were randomized. The order of this block-wise randomization was kept for each cycle. Each localization was measured three times, nonconsecutively. All measurements were spread over 2 days.
Subects Ten healthy volunteers (6 male, 4 female; median age = 29 years; range = 23–67 years), predominantly students, were included in the study after proving eligibility in an anamnestic and physical internal and orienting neurological examination. We screened for illnesses that could present a risk for test subjects or alter the results of the measurements. These include psychiatric, neurological, and chronic pain illnesses, as well as movement and coagulation disorders. A tendency to bleed was excluded with a validated questionnaire (Koscielny et al., 2004). None of the subjects had a recent history of musculoskeletal pain or took pain medication. One female was on an oral contraceptive. All 4 females were not having
Behav Res
Head and Neck
Trunk
Arm
Hand
Leg
Fig. 4 Overview of all 29 measurement sites. ndom = non-dominant side; dom = dominant side
their period. A male subject took finasterid because of a prostate hypertrophy but did not report any side effects. All volunteers were right-handed pursuant to the Edinburgh Handedness Inventory (Oldfield 1971). None of the subjects were obese (mean body mass index [BMI], 22.9 kg/m2; range, 19.4–27.5 kg/m2). One subject was a smoker. The study was approved by the local ethics committee and was conducted in accordance with the declaration of Helsinki and its amendments. Informed written consent was obtained from all volunteers prior to the inclusion of the study. All subjects received a monetary compensation for their participation. Pain perception and self-assessment of subjects After each measurement test, subjects were asked to place a mark on a 100-mm visual analog scale (VAS) to indicate the subjective perception of pain intensity at the measurement site ranging from 0 mm = no pain to 100 mm = most pain imaginable. The State Anxiety Inventory (STAI-S; Spielberger, 1983) was used to measure possible experimental anxiety. The
inventory is easy to score and possesses good validity and internal reliability (Cronbach’s α=.91) (Spielberger, 1983). The Brief Symptom Inventory (BSI) is a 53-item selfreport questionnaire designed to reflect the current psychological symptom status (Derogatis, 1975). Each item is rated on a 5-point Likert-type scale ranging from not at all to extremely. The items define a broad spectrum of perceived restrictions relative to physical and psychological symptoms that occurred in the preceding 7-day period. Symptoms are assigned to nine subscales, which represent domains of psychopathology (e.g., trait-anxiety, depression, and somatization). The Karolinska Sleepiness Scale (KSS) was used for evaluating possible subjective sleepiness, using a 9-point scale, ranging from extremely alert to extremely sleepy/fighting sleep (Åkerstedt & Gillberg, 1990). Experimental procedure The first day began with a medical examination to check for possible pain-related disorders and current illnesses (Fig. 3).
Behav Res
Afterward, subjects filled out the BSI and were introduced to the experiment. Initially, three test measurements were conducted using a conventional hand-held pressure algometer (Sauter FA 500, Greenwich, U.S.), followed by seven test measurements with the automatic algometer (Fig. 1). The primary goal was to approach the concept of pain thresholds and to reduce discomfort and possible anxiety. The STAI-S was filled out four times: (1) during the medical examination, (2) after the algometer room was entered for the first time, (3) before the first experimental measurement, and (4) after the last experimental measurement. Under certain circumstances, the investigator’s gender can influence the reaction of the subject during the experiment (Levine & De Simone, 1991). This is why the investigator was always of the same sex as the subject. Interpersonal expectancy effects (Rosenthal & Rubin, 1978) were minimized by means of neutral and standardized instructions and by practicing the algometric procedure. The measurements took place the next day and were spread over 2 experimental days. The average temperature was 23.1 °C (SD = 0.81), and the relative air humidity was 36.9 % (SD=6.6). Subjects were blindfolded for each measurement. Pressure was applied perpendicular to the skin. The PPT was signaled by pressing or releasing the response key and was defined by the following instruction: “Please press or let go, when pressure transitions into beginning pain.” The increase of pressure was set at 5 N/s, except for the chewing muscle (M. masseter ndom) and temple (Os temporale ndom), which were set at 2 N/s owing to low PPTs at those sites, which were obtained during our preliminary experiments. After each measurement, subjects immediately recorded their perceived level of pain using the VAS. Subjects were also asked to report their personal observations (e.g., unusual sensations, lingering pain, etc.). No lingering pain was reported. If the pressure was not distributed evenly on the skin, the subject was asked to identify the exact point where possible edges or other special features were felt. Those measurements were repeated at the end of all regular measurements. Breaks were taken on the basis of the subject’s individual well-being and his or her vigilance. Statistical methods Due to the small sample size of this pilot study, data are evaluated exploratively and are presented descriptively. According to an analysis of distribution, PPTs (and pain intensity ratings) are log-normal distributed. As a measure of central tendency, we used the mean value of three measurements for single localizations of every subject (Hooten, Rosenberg, Eldrige, & Qu, 2013; Zolnoun et al., 2012). The geometric mean was used for aggregating the values of all subjects and/or measurement sites. In order to quantify how the variation of the PPT, as well as the rating of pain intensity (VAS), was distributed over the
factors subject, measurement localization, and random error, we fitted a hierarchical linear random effects model. Spearman’s rank correlation coefficient with the corresponding 95 % confidence interval (95 %-CI) was used for the analysis of bivariate correlations, and the Friedman test was used to compare repeated measurements. All presented p values are two-sided. Boxplots and an illustration of normalized geometrical means with corresponding confidence intervals were used for the graphical presentation. The analysis was performed using the statistical software SAS version 9.3 (SAS, Cary, NC), SPSS version 20 (IBM, Armonk, NY), and R version 2.12.2 (R Development Core Team).
Results Functioning and safety of test facility The complete test facility has proven to be reliable, easy to operate, and durable in all phases of the design and testing process. During the pilot study, no major technical disturbances occurred in the operation of the automatic pressure algometer. The increase in pressure was, for almost all measurements, quasi-linear (Fig. 5). Single fluctuations of force were observed as a result of different tissue properties or movement caused by breathing. Including our preliminary tests, we conducted roughly 1,500 PPT measurements and found no damaging effects on the skin. Even after repeated measurements at various sites, neither skin irritations nor local hematomas were observed. During our preliminary tests, we discovered that measuring 29 sites is very time consuming. Despite using adjustable fixation devices, one cycle lasted up to 5.5 h. The more efficient approach as described for the pilot study of grouping measurement sites into blocks of neighboring localizations reduced each cycle, on average, by 2 h. The mean time of 1 experimental day was 5 h and 19 min (SD=24 min). By minimizing the amount of reconstructions of our fixation devices, we saved 6 h per subject, on average, for the complete experiment. The vacuum cushions adapted perfectly to all anatomical structures. All fixation devices were ergonomic and comfortable, according to the direct feedback of our volunteers. One subject described the measurements at the kneecap and another at the temple as unpleasant. Four out of 10 subjects reported some difficulties in defining the onset of pain during the test measurements but were able to make correct assessments during the experiment. Pressure pain thresholds The average PPTs at 29 measurement sites varied among the subjects (Fig. 6). With the exception of the forehead, low PPTs
Behav Res
were predominantly recorded at or near cranial sites, which include the chewing muscle, the temple, and the neck muscle. The highest PPTs were measured at the palm of the hand, the index finger tip, and the calf muscle.
[N]
100
Pain perception and self-assessment of subjects
80
60
40
20
0 0
5
10
15
20
25
[s]
Fig. 5 Force-time curve of a measurement at the calf muscle (ndom)
Fig. 6 Pressure pain thresholds (PPTs) of all localizations. PPTs of 10 test subjects were averaged over three repeated measurements. Boxplot: The 1st quartile value is at the lower horizontal line of the box; the 3rd quartile value is at the upper horizontal line; the thick line in the box represents the median. The whiskers (the lines that extend out the top and bottom of the box) represent the highest and lowest values that are not outliers. Outliers (values that are between 1.5 and 3 times the interquartile range) are represented by circles beyond the whiskers. The breast muscle was measured only with male subjects (n=6)
The median pain intensity score after measuring the PPT was 6 mm (min = 0 mm, 2nd quartile = 3 mm, 3rd quartile = 11 mm, max = 32 mm). The maximum rating of 32 mm was recorded after a measurement of the palm of the hand on the nondominant side (54.9 N). Figure 7 shows a tendency for higher pain intensity ratings at more sensitive measurement sites (e.g., chewing muscle) and vice versa (p=.051, effect of VAS in a random slope mixed model with VAS as the only fixed effect). This was no longer the case after accounting for the measurement sites (p=.26). Within all 29 measurement sites, only the pain intensity rating of the palm of the hand (ndom) showed an inverse relationship with the corresponding PPT (Spearman’s rho = −.37, 95 %-CI [−0.64; −0.24], p=.046). On the basis of the observations of all experimental days, the median score of the KSS was 3 (min = 2, 1st quartile = 3, 3rd quartile = 4, max = 6), indicating an acceptable level of vigilance. Scores of 6 (between neither awake nor tired and tired but no problems staying awake) were reported twice. Scores decreased in both cases after having taken a break. State-anxiety scores of our subjects were generally low. The overall level of state-anxiety changed only marginally on a group basis during the four different time points, χ 2 (3) = 1.16, p = .76 (Friedman test) (Fig. 8). Two subjects revealed slightly higher scores, one of which exceeded the postulated level of 50 during the medical examination, indicating an elevated stateanxiety level (Hermes, Matthes, & Saka, 2007), and the other after entering the algometry room for the first time. In both cases, state-anxiety normalized on the second day before the first round of measurements took place. The scores of the trait-anxiety subscale of the BSI correlated negatively with the aggregated geometrical mean PPT score of all measurement sites (Spearman’s rho = −0.71, 95 %CI [−0.94;−0.23], p
