Identifying pain in children with CHARGE syndrome

2.1 Participants

Parents of children with CHARGE syndrome were recruited from the CHARGE Syndrome Foundation research listing. Participants were recruited if they had a child with CHARGE 18 years of age or under. The Institutional Review Board approved the study and informed consent was obtained from each subject.

Fifty-three parents participated in this study, 81.1% (n=43) were mothers. Male children comprised 64.2% (n=34) of the sample and the mean age of children with CHARGE was 8.87 years (SD=5.00), with a range of 1–18 years of age. All participants’ children had a clinical diagnosis of CHARGE. In addition, 44.2% (n=23) of children had been genetically tested for the CHD7 mutation, which is known to cause CHARGE [13]. Twenty-nine children had not been tested and one participant did not know if their child had been tested. Of the children tested for the CHD7 mutation, 65.2% (n=15) tested positive, as would be expected based on current known rates [26]. Additionally, all of the children had at least one major surgery, ranging from 1 to 47 procedures, with a mean of 11.85 surgeries (SD=9.82).

Children in this study had significant vision and hearing problems despite use of corrective aids (e.g. eyeglasses, hearing aids, or other hearing devices). Approximately 80% of the sample displayed a range of vision impairments with correction (e.g. eyeglasses) from “some trouble seeing” to “complete blindness.” Even more significantly impacted was hearing with approximately 90% of these children with CHARGE displaying a range between “some hearing deficits” with correction (e.g. hearing aids, Cochlear implant, etc.) to “complete Deafness.”

Fifty-five percent of individuals with CHARGE in this study were able to use some form of communication (e.g. vocal strategies, sign language, or pointing to a specific body part) to communicate discomfort. Of those with some form of communication, 34% (n=10) of participants indicated the child could communicate acute pain, but could not differentiate or describe chronic pain. According to participant ratings on the Non-Communicating Children’s Pain Checklist-Revised, approximately 39% of individuals with CHARGE could make a specific sound or word for pain a number of times when in pain or continuously (e.g. a word, cry, or type of laugh), while approximately 9% of the children rarely displayed a specific sound or word for pain. Approximately 52% of these individuals with CHARGE had no sound or word for pain.

2.2 Measures

2.3 Procedures

An introduction letter, consent form, measures, and detailed procedures and directions for completing the measures were mailed to each participant through the United States Postal Service with a stamped return addressed envelope.

Participants first completed the Pain in CHARGE Syndrome Demographics Sheet. Second, parents completed the PPP and NCCPC-R, based on observations of their child’s behavior when the child was suspected to not be in pain. Next, participants completed the PPP and NCCPC-R during their child’s most troublesome pain, e.g. gastrointestinal discomfort, or when pain was suspected. Participants were directed to review items on both instruments prior to daylong observations of their child. The PPP and NCCPC-R (no pain and when pain was suspected) were then completed at the end of each day (no pain day and pain day). As previously stated, it should be noted that a daylong observation is not standard procedure for completing the NCCPC-R. Standard procedure would have parents observe their child for a 2-h duration rather than across the day. If the observation took place across the day, however, it was hoped that all items would be included in the observation, e.g. eating and sleeping, rather than completing the NCCPC-R after a 2-h time period when all items may not have been observed. Additionally, the PPP is a daily measure of pain; therefore, to avoid confusion, the decision was made to change the 2-h observation of the NCCPC-R to a daily observation, in congruence with the PPP.

Lastly, participants completed the Pain in CHARGE Syndrome Questionnaire regarding specific questions about their child’s pain experiences. This questionnaire was completed last as it was necessary to complete Section II of the Pain in CHARGE Syndrome Questionnaire once the caregivers had rated their child’s behavior on the PPP and NCCPC-R. Section II addressed questions regarding how these instruments were able to help participants determine a difference in their child’s behavior when in pain and questions on social validity.

Participants then returned all measures including the original consent form in the self-addressed, stamped return envelope to the Research Laboratory.

3.1 Pain experience

All participants (n=53) completed the Pain in CHARGE Syndrome Questionnaire, which asked if parents were able to determine when their child with CHARGE was experiencing pain and the estimated amount of pain their child experiences. Just over 75% (75.5%, n=40) of the sample indicated they could determine when their child with CHARGE was in pain while 11.3% (n=6) were unsure. The results indicate a large majority of parents have at least one strategy to determine when their child is in pain, which is particularly important given approximately 45–52% of the sample was unable to vocalize their discomfort or pain experiences. Approximately one-fourth of parents surveyed, however, were not able to determine when their child was experiencing pain or discomfort. This further illustrates the need to provide parents/guardians with strategies they can use to determine their child’s pain experiences, particularly when the child is unable to communicate their acute or chronic pain experiences. An independent samples t-test was conducted to examine if the 25% of parents who could not determine their child’s pain varied by the child’s age, vision, or hearing and the results were not significant.

Nearly all participants (94.3%, n=50) indicated their child with CHARGE experiences pain. Participants estimated their child’s pain experiences on the following scale: no pain, very little pain, little pain, occasional pain, very frequent pain, and always in pain. Based on participant ratings, approximately 94% of the children with CHARGE in this study experience pain of some degree with the greatest frequencies being “very little pain” and “occasional pain,” representing a bimodal distribution. The following ratings were provided: no pain (n=3, 5.7%), very little pain (n=17, 32.1%), little pain (n=3, 5.7%), occasional pain (n=20, 37.7%), very frequent pain (n=9, 17%), and always in pain (n=1, 1.9%). More than half of the children (56.6%) were rated as having occasional pain or greater.

Additionally, the PPP measured how often children with CHARGE have pain free days based on the following four-point rating: all the time, most of the time, some of the time, and hardly ever. The modal rating was most of the time, as was the mean rating (M=2.17, SD=0.79), based on the results shown in Table 1. Further, parents were asked if their child has pain even when they are having a “good day,” based on the following five-point rating: no pain, mild pain, moderate pain, severe pain, and very severe pain. The results are presented in Table 1. Parents did not indicate severe or very severe pain on a “good day” and a bimodal distribution of no pain and mild pain was found. This indicates that for just over half of the children with CHARGE in this study (52.8%, n=27) mild or moderate pain occurs even on a good day, whereas the remainder of parents would describe their child’s good days as pain free.

In order to better understand non-vocal pain identification, participants were asked to list the signs and behaviors they use to determine when their child is experiencing pain. Results were placed into the following categories from the NCCPC-R: vocal, social, facial, activity, body and limbs, physiological, eating/sleeping, and an additional category was added that included behavioral challenges, self-injurious behaviors, and dangerous behaviors. Results are listed in Table 2.

Further, the Pain in CHARGE Syndrome Questionnaire sought specific information about what caused pain for their child with CHARGE. Table 3 details the hierarchy of the most intense painful experiences and average duration of common CHARGE characteristics that are suspected to cause pain. Participants provided an estimated number of days per year their child with CHARGE exhibited pain for each category and rated their child’s pain intensity on a Likert scale from 0 to 4 (0-no hurt, 1-hurts little bit, 2-hurts more, 3-hurts whole lot, and 4-hurts worst) on a facial expression chart, the Faces Rating Scale [31].

The Spearman Rho statistic was used to compare the paired observations of duration of pain and intensity of pain. The mean duration of pain was ranked and the mean intensity rating was also ranked from 1 to 13 for each of the CHARGE characteristics. A difference between the ranks was determined and squared and a rank difference coefficient of correlation (Spearmans Rho Coefficient) was calculated. It should be noted that there were no tied ranks.

A moderate negative correlation between the mean intensity of pain and the mean duration (days of pain per year) of pain among individuals with CHARGE in this study was found, ρ=−0.34. There was a tendency for intensity of pain to increase for those sources of pain that were of shorter duration. Thus, individuals with CHARGE were rated to have more intense pain during periods of acute pain (shorter duration) than during periods of chronic pain (long-term pain).

According to the report of these parents, the most intense pain for children with CHARGE is migraines, as indicated by the highest mean intensity and by rankings of “hurts more” to “hurts worst.” The most chronic characteristics (appearing 95 days or more each year) were: gastroesophageal reflux, difficulty swallowing, breathing, hip/back pain, abdominal migraines, and muscle pain. When looking at the range of pain intensity, those with the “hurts worst” rating are breathing, difficulty swallowing, chronic recurrent otitis media, sinusitis, migraine, abdominal migraine, constipation, gastroesophageal reflux, hip/back pain, and surgery pain; therefore, coughing, jaw discomfort, and muscle pain appear to produce pain but would not be described as the worst pain, according to this sample.

Participants scored the items on the NCCPC-R and PPP for when their child was not in pain, and again when their child was in pain. The expectation was that there would be considerable difference in the ratings on the items for the two administrations because ratings at baseline should be lower than ratings during pain, as the scales have been developed to measure painful behaviors. As a result, higher ratings would occur during pain. This serves to re-evaluate the validity of the NCCPC-R and PPP with a unique population of CHARGE syndrome, in which individuals experience communication deficits combined with vision and hearing loss. To evaluate whether parents’ ratings for pain were greater than ratings for when the child was suspected to not have pain on the NCCPC-R, a paired samples t-test was conducted. On the NCCPC-R, the mean rating on the items when the child was in pain was 61.44 (SD=19.18) and the ratings for when they were not in pain was 38.93 (SD=7.91). The difference between these ratings was significant beyond the 0.05 level, t (40)=8.15, p=0.000, 95% CI [16.93, 28.10]. The standardized effect size was d=1.3, a large effect.

Similarly, a paired samples t-test was conducted to evaluate whether participants’ ratings for pain were greater than ratings for when the child did not have pain on the PPP. The mean pain ratings (M=25.62, SD=13.06) were significantly greater than the mean ratings for no pain (M=10.77, SD=8.09) with significance beyond the 0.05 level, t (51)=9.59, p=0.000, CI 95% [11.74, 17.96]. The standardized effect size was d=1.3, a large effect.

On both instruments, all of the items were rated lower when not in pain than when in pain. To further extend our understanding of pain behaviors for individuals with CHARGE, it is necessary to determine if there are specific behaviors that indicate pain more than others. Baseline ratings (no-pain observations) were compared to pain observations and several behaviors were shown to have at least a one-point difference (NCCPC-R range of 0–4 and PPP range of 0–3), as can be seen in Table 4.

Items that had at least a one-point mean difference from NCCPC-R No Pain to NCCPC-R Pain were: not moving/less active/quiet (#14), tears (#25), not cooperating/cranky/irritable/unhappy (#5), crying (#2), moaning/whining/whimpering (#1), less interaction with others/withdrawn (#6), turning down of mouth/not smiling (#11), being difficult to distract/not able to satisfy or pacify (#8), and a furrowed brow (#9). Items with at least a one-point difference on the PPP are very similar to items from the NCCPC-R: cheerful (#1), cries/moans/screams/whimpers (#4), is sociable/responsive (#2), grimaces/screws up face/screws up eyes (#9), frowns/has furrowed brow/looks worried (#10), and hard to console/comfort (#5). These items appear to have the greatest difference in presentation from no pain and are likely behaviors many individuals with CHARGE would display when experiencing discomfort or pain. Table 4 presents in hierarchical order the mean difference score on the NCCPC-R and PPP from no-pain to pain.

Several behaviors appeared not to display a meaningful difference between pain and no pain across both measures including: flexing inward/drawing legs up (PPP #15), stereotypical movements/jumping/seizures (PPP #20), flopping (NCCPC-R #16), shivering (NCCPC-R #22), and jumping around/agitation/fidgety (NCCPC-R #15). Such behaviors may not be useful indicators of pain for use with this population due to their common presentation in CHARGE. A Pearson correlation was conducted between the pain measures of the NCCPC-R and PPP to determine the direction and degree of the relationship between the two pain total score variables. A significant positive relationship was found, r=0.67, p=0.00 indicating there is a strong association between the NCCPC-R and PPP and that both instruments measure the same construct when both are completed at the same time. Further, this indicates that high scores on the NCCPC-R tend to be associated with high scores on the PPP for individuals with CHARGE.

The NCCPC-R and PPP have several similar items and some items on the NCCPC-R appear to be combined on the PPP, e.g. crying, screaming, whimpering. Each measure also has items that are not included on the other, e.g. the PPP contains self-harming whereas the NCCPC-R does not. As previously shown in Table 4, many of the items that produced the largest difference scores for individuals with CHARGE were similar across both the NCCPC-R and PPP. Items with the greatest difference score (0.73 or above) that are similar across both measures, NCCPC-R and PPP, respectively, included: not cooperating vs. is cheerful, crying vs. cries/moans/screams/whimpers, less interaction with others vs. appears withdrawn/depressed, furrowed brow vs. frowns/furrowed brow/looks worried, difficult to distract/not able to satisfy or pacify vs. hard to console/comfort, eating less vs. reluctant to eat, not moving/less active/quiet vs. resists being moved, change in eyes vs. grimaces/screws up face/eyes. Other similar items with at least a half point difference score on both measures include tensing, stiffening, and spastic movements, and touching or rubbing parts of the body that hurt.

When comparing the NCCPC-R and PPP, it was of interest to determine which of the two instruments is a better measure of pain in CHARGE. Cronbach’s alpha is a measure to determine how consistent or closely related items are in a measure. The PPP (n=52) and NCCPC-R (n=41) had the same alpha coefficients of 0.92 indicating very good internal consistency. Although, this is the only measure of reliability available in this study, it would appear that both are reliable measures, when based on the parents’ assessments of their child’s presentation.

Although both measures appear to be internally consistent with this population, 55% of parents (n=18) stated the PPP best described their child’s pain, while 45% of parents (n=15) believed the NCCPC-R was a better measure. Additionally, more parents (n=11) completed the PPP but chose to not complete the NCCPC-R. It does appear, however, that many parents had difficulty determining which assessment was a better measure of their child’s pain because 20 parents either left this item blank or indicated both measures were helpful.

The majority of parents (79%) stated that the non-vocal multidimensional pain measurements helped them to estimate the type or amount of pain their child was experiencing, whereas 2% felt the NCCPC-R and PPP “may be helpful.” Additionally, 83% of parents largely agreed that the use of non-vocal pain assessments would be beneficial in the child’s classroom, employment setting, etc. Ten-percent of parents felt the use of such instruments “may be helpful.” Several parents suggested the NCCPC-R and PPP would be useful in healthcare settings and for their child’s medical professionals.

4.1 Limitations and future directions

This investigation has several limitations. First, the results rely on parent observation and report. It is unknown from this investigation whether others, particularly individuals not familiar with the child (e.g. nurses, physicians, new care staff), would observe and rate the child’s behavior in the same fashion as parents or others who know the child well.

A second limitation is that this study was restricted to individuals with CHARGE aged 18 years or younger. Pain may present differently across the lifespan as communication strategies improve or as the nature of the pain experience changes, with some pain subsiding, and other pain developing. For example, Blake and colleagues found that there are medical conditions, such as migraines, that appear to increase during adolescence [37]. Third, as previously indicated, the non-vocal instruments used were shown to be valuable to identify pain in this revalidation study for CHARGE; however, an instrument designed specifically for CHARGE may be more valuable with all items contributing as indicators of pain. This investigation included some children who could, in some fashion, indicate pain, which may present as a limitation. By including individuals who can self-report their pain experiences, the results may be confounded by the child’s vocal expression of pain rather than by behavioral expressions. Conversely, the child’s ability to vocally communicate may be beneficial to parents/caregivers in confirming the behavioral expression they believe to be indicators of pain truly are pain indicators.

An investigation to understand the relationship between acute and chronic pain is needed and how each impacts adaptive, academic, and overall functioning. It would also be useful to investigate the relationship of chronic pain to pain sensitivity or tolerance for pain. Interestingly, this investigation also found some areas of adaptive functioning were reduced during episodes of pain. Communication is an important area of adaptive functioning that may be impacted by pain. This relationship needs to be explored more closely for individuals with CHARGE, including how a reduction in communication and an increase in pain may produce challenging behavior, as well as how vision and hearing loss may impact pain and adaptive behavior.

In conclusion, Bottos and Chambers stress that considerable research needs to be conducted to identify pain among individuals with particular disabilities and the specific types or sources of pain these individuals experience [1]. For individuals with CHARGE syndrome, there may be many unique sources of pain given the nature of their syndrome. However, the actual identification of pain non-vocally can be difficult, as the pain induced behavior they display may not be typical for other individuals with developmental disabilities or genetic syndromes, making non-vocal scales designed for these groups not as valid in CHARGE. This was shown in the number of items on the PPP and NCCPC-R that did not display a meaningful difference between baseline and pain ratings. Further, participants identified multiple behaviors, particularly challenging behaviors, as indicators for painful experiences; however, these items are not covered on the instruments investigated. As a result, additional investigations are needed to develop a non-vocal pain assessment for individuals with CHARGE that includes challenging behaviors and significant items from this investigation.