Pain in adults with cerebral palsy: impact and solutions

Authors


  • CONFLICTS OF INTEREST
    The author declares no conflicts of interest.

Laura K Vogtle at Department of Occupational Therapy, University of Alabama at Birmingham. Birmingham, AL, USA. E-mail: lvogtle@uab.edu

Abstract

Studies of health in adults with cerebral palsy (CP) have identified pain as a significant concern. Investigations regarding incidence, intensity, and location in adults with CP found that increasing age and inactivity appeared to be related to pain. Activity and participation in adults with CP seem to be only moderately affected by presence of pain. Various sources of pain have been identified in adults with CP but have not been well studied. These include orthopedic issues, poor bone mineral density and related fractures, dental and jaw problems, and nutrition-related pain. Limited healthcare utilization studies suggest that adults with CP use healthcare services, especially preventative and rehabilitative services less, and do not consult physicians regarding pain. Medication solutions for chronic pain are not well studied. Botulinum toxin and intrathecal baclofen have been demonstrated to minimize pain; however, the impact of other medications needs further investigation. Other interventions for pain include small studies examining the use of biofeedback and exercise. Larger studies are needed to establish effectiveness. In order to prevent future generations of adults with CP from experiencing high levels of pain, environmental sources of pain need more specific study, as do interventions that are affordable and easily accessed.

LIST OF ABBREVIATIONS
AED

Anti-epileptic drugs

BMD

Bone mineral density

BoNT-A

Botulinum toxin-A

ITB

Intrathecal baclofen

Cerebral palsy (CP), defined as ‘a group of permanent disorders of the development of movement and posture causing activity limitation that are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain,’ has been associated with pain in both children and adults.1 Some pain results from the movement impairments that characterize CP, most notably musculoskeletal factors, including soft-tissue limitations, joint deformity, and spasticity. However, recent studies have suggested that the scope of pain issues in this population is more complex than musculoskeletal concerns alone, particularly in adults. Muscle weakness, fatigue, deterioration of functional status with aging, limited access to healthcare services, falls, fractures, health conditions not related to CP, and environmental factors may play a role in pain experienced by adults. This paper will discuss studies relating to pain in adults with CP, the impact of pain on activity and function, and interventions that have been demonstrated to alleviate pain.

Pain in children with CP

A limited number of studies on pain in children with CP exist.2–5 The subjective nature of pain, combined with the challenges assessing it in a diverse population where communication difficulties and cognitive impairments are common, make the study of pain in children with CP complex. Parent responses may be utilized in lieu of child reports, sometimes with conflicting outcomes.4,6 Houlihan et al.4 reported that 39% of the children in their sample were reported by parents to have pain ranging from a few times a month to almost every day, which was significantly higher than the norming sample for the assessment used. In contrast, Kennes et al.6 found that 86.5% of the children in their sample were reported by parents to be pain-free. Oberlander7 suggested that the relationship between pain in CP and upper-motor-neuron lesions remains to be described. Multiple sources of pain, including dental and orthopedic pain, positioning issues, feeding problems, reflux, constipation, stretching, and bracing, have been described in the literature.3,4,8

In spite of methodological concerns in existing studies, the consensus appears to support the presence of pain in a high percentage of children with CP. Given such findings, it should be no surprise that adults with CP report significant pain as well.

Pain in adults with CP

Over the past 15 years, chronic pain in adults with CP has been identified as a significant concern. Studies examining health status in adults with CP were the first to report a high incidence of pain in this population.9–11 More recent investigations, focused on pain intensity, location, and number of pain sites across samples, support these findings.12–14 In both their first and second cohorts, Schwartz et al.14 and Engel et al.12 found that 67% of their sample had one or more chronic pain sites. The back, hip, and lower extremities were the most common pain locations in both studies. The sample in Schwartz et al.14 comprised 84% adults with quadriplegia as compared to 74% in the Engel et al.12 sample; although individuals in the former sample had greater impairment, those in the latter sample had greater pain intensity. Increased pain with age was reported in the first study, but not mentioned in Engel et al.12 Jensen et al.15 later followed a sample of 50 adults with CP over 2 years and found that pain was stable over time.

Jahnsen et al.13 generated survey responses from 406 adults with CP. Distribution of CP was broader than in studies by Schwartz et al.14 and Engel et al.,12 with 38% of participants having hemiplegia, 36% diplegia, 17% dyskinesia, and 8% quadriplegia. Eighty-two percent of these adults reported pain in one or more sites, and significantly more women than men reported pain, a finding that is consistent with pain studies in the general population. Increased pain was reported with increasing age, again replicating results from earlier studies. The back pain was the most frequently reported pain site; however, pain location in other body areas varied by distribution of CP. A broader sample of CP distribution (35% diplegic, 21% hemplegic, 12% quadriplegic, 22% dyskinetic, and 9% mixed) was investigated by Andersson and Mattsoon;16 79% of this sample reported pain primarily in the back, shoulders, and hips.

The development of the Gross Motor Function Classification System (GMFCS)17 has facilitated understanding of the presence of pain across types and distributions of CP. Two studies have investigated the relationship between pain and GMFCS levels. Fifty-nine percent of adults with CP in a Dutch study reported pain at locations similar to those reported in previous studies (hip, leg, and back most frequently) and no association between pain and GMFCS level.18 Sandstrom et al.19 assessed the relationship between GMFCS levels and various aspects of health and function. Pain was reported in 68.75% of their sample, with no correlation between GMFCS level, pain location, and pain intensity. A trend between inactivity and pain was noted.

Coping with pain in CP

Cognitive-behavioral models of pain and coping suggest that the use of certain coping strategies predicts better adjustment to chronic pain. Engel et al.20 investigated the use of coping strategies in 50 adults with CP who reported using such strategies more frequently than other populations with chronic pain. These strategies included task persistence, attention diversion, reinterpretation of pain sensations, and prayer/hope. A later longitudinal study revealed that increasing more positive coping strategies over time resulted in increases in function. Less positive strategies, such as ‘catastrophizing’, were related to decreases in function.

A body of evidence documents the presence of pain in adults with CP. Although the evidence describing increased pain in persons with greater motor impairments is more substantial, there can be no question that adults with milder levels of motor impairment also experience significant pain. One limitation across studies is that only adults who are cognitively able to respond to questions and surveys are included in samples. Methodologically this makes sense, but it does eliminate information on pain status in persons with CP who have cognitive impairments. Understanding pain incidence for adults who are cognitively limited may alter care provision and improve comfort.

A second group that is not well represented in pain studies is adults with dyskinesia. Since persons with this type of CP make up a small percentage of the CP population, the small samples are logical. However, sources of pain in individuals with dyskinesia may differ from those in persons with spastic CP, since the two movement disorders vary considerably.

Sources of pain

The primary, initial disorder in CP is non-progressive brain damage. That damage results in motor impairments and subsequent effects on body structures and functions (muscles, tendons, ligaments, joints, posture, motor control). These effects are called secondary conditions, defined by Kenne et al.21 as ‘preventable physical mental and social disorders resulting directly or indirectly from an initial disabling condition.’ Some secondary conditions resulting in pain in adults with CP are described below.

Orthopaedic pain

The back, hips, and lower extremities have been identified as consistent sources of pain in virtually every investigation on pain in adults with CP. Given the occurrence of subluxing and dislocated hips; abnormalities of the foot, knee, and ankle; pelvic obliquity; scoliosis; and contractures, particularly in adults with more significant motor impairment, these outcomes are not surprising. Studies of non-ambulatory adults with CP found significant associations between deformity, asymmetry, and, in some studies, presence of osteoarthritis and pain.22,23 Of the individuals reported in these two studies, 30% and 47% respectively, had pain in hip-loading situations, defined as putting on trousers, being lifted from bed, lying in bed, sitting in a wheelchair for extended periods, and undergoing physical therapy in the lower limbs. Twenty out of 160 adults with CP reported by Boldingh et al.22 had unsuccessful surgical interventions, resulting in long-term hip pain. Hodgkinson et al.23 found that only 13.6% of all participants received medical intervention for their pain. In this study, in contrast to other studies, younger participants reported more pain than older persons. Cooperman et al.24 reported that more than half of their sample, followed for a mean time of 18 years, experienced pain following dislocated hips. Only a small percentage of the sample had successful hip reductions. In the only long-term follow-up study on foot surgery, a significant incidence of chronic pain was reported in participants with triple arthrodeses.25

All these studies except that of Tenuta et al.25 assessed outcomes in persons who were non-ambulatory and had severe motor impairments. Osteoarthritis was reported as a cause of pain in several studies,23,25 while other studies noted that pain was related to positioning and care activities. Consistent with the work of Schwartz et al.14 and Engel et al.,12 most of these studies included adults with significant motor impairment. However, Jahnsen et al.26 studied decreases in locomotor skills with aging and included a better representation of the distribution of impairment associated with CP, including the ambulatory population. Pain in the lower extremities was one reason adults indicated they had stopped walking.

Pain associated with low bone mineral density and fractures

Low bone mineral density (BMD) and fractures have been documented in children and adults with CP.27–29 These fractures have been primarily found in persons who are non-ambulatory, demonstrate poor nutrition and growth, and take anti-epileptic drugs (AEDs), which have been noted to increase the risk of osteoporosis.30 There have been virtually no studies regarding the effect of menopause on the development of osteoporosis in women with CP. The relationship between pain and BMD is an especially important area because many adults with CP who walked as children cease to do so in early adulthood, thereby increasing the risk for osteoporosis.9,31,32

Dental pain

Dental pain has been reported in association with CP9,11,33 Traumatic injuries to teeth have been reported in children with CP, as have increased risk of joint and muscle problems in the orofacial area; however, no corresponding studies exist in adults.34–36 Seizure disorders are reported in 30% of persons with CP, and use of some AEDs has been noted to cause difficulties with oral hygiene.37

Dental care is mandated by Medicaid for children who qualify and is the primary source of funding for dental care in children with developmental disabilities. However, Medicaid reimbursement for dental care in adults with developmental disabilities is an elective Medicaid service and in many states is limited to extractions only.38 These findings indicate that the lack of dental care in adults with CP noted in earlier studies is at least partly due to lack of funding for dental services. Another factor, discussed by Waldman et al.,38 is that dental residents do not receive appropriate training in the treatment of persons with developmental disabilities. This lack of training may have an impact on their willingness to treat this population.

Nutrition-related pain

Growth failure has been consistently noted in children with CP, particularly in those with severe motor impairments. In a multi-site study, parents reported high incidence of pain associated with gastrostomy tubes in children.4 Nolan et al.39 discussed pain in malnourished children that was associated with prolonged pressure from sitting on boney prominences. The same concerns exist for adults with CP who are underweight, especially given the infrequent change of position often seen in this population due to difficulties with securing personal care assistants and challenges associated with moving a non-ambulatory adult. Another source of pain and discomfort in persons with CP is constipation and vomiting related to feeding problems and poor nutrition.4,39

While underweight is the most common nutritional concern mentioned in association with CP, obesity also exists among this population and its incidence is growing. Recent studies in children with CP have found increases in the prevalence of obesity, particularly in children at GMFCS levels I and II.40,41 Obesity in childhood has been significantly associated with obesity as an adult, a factor that should raise concerns for the CP community.42–44

Cathels and Reddihough33 found that 8% of the 149 adults in their survey reported obesity, resulting in exacerbations of arthritis and difficulty with independent mobility. Two-thirds of the adult females with CP in a study by Van Der Sloot et al.45 were obese. No other specific studies on obesity in adults with CP exist; however, several articles that note obesity incidence increases in adults with intellectual impairments have included subsets of adults with CP.46–48 Obesity is associated with increases in pain, decreased self-care, and lost productivity in typical adults.49,50 The potential impact on adults with CP, who already face significant issues with pain, as well as limitations in physical activity, is a matter of serious concern.

Utilization of healthcare services by adults with CP

Medicaid is a major source of funding for healthcare of adults with CP in the USA. Funding restrictions, as well as lack of means of transportation to service providers and accessible healthcare facilities, have a negative impact on the use of these services. There is limited evidence suggesting that healthcare utilization in adults with CP decreases after age 21. A significant decline in utilization across all healthcare services in adults with CP between the ages of 18 and 36 has been described in the Netherlands.18 A lack of preventive services for adults with developmental disabilities (including CP) living independently, with family or friends, or in community facilities, has been described in the USA.48 High rates of obesity, lack of dental care, and increased polypharmacy for psychiatric and epileptic conditions were reported in this study. Fewer than 50% of women with CP received regular Pap smears. An investigation regarding the use of physical rehabilitation services in a sample of women with multiple sclerosis, CP, and spinal cord injury found that all these individuals had difficulty accessing rehabilitation services. Persons with low income or less education, and persons who were married, had the most problems with access.51

In contrast to the previous studies, a Canadian investigation found hospital admission rates for adults with disabilities were seven times higher than those in the general population.52 Compared to children with CP, adults with CP had a greater number of outpatient visits and emergency room visits, but fewer specialty physician visits. The most frequent reasons for admission to the hospital were respiratory problems, digestive problems, injury, and poisoning.

Overall, these studies suggest that adults with disabilities, including CP, have difficulty accessing both preventive healthcare and rehabilitation services. Clearly more study is needed on healthcare utilization, adults with CP, and other health indicators in order to understand the role of healthcare utilization in identifying sources of pain and pain management.

Environmental factors

The International Classification of Functioning, Disability and Health53 emphasizes the complexity of factors related to aspects of disability. These factors include aspects of the physical environment as contributors to body structure/body function, disability, and participation. Elements of the physical environment include accessibility for persons with disabilities, weather, terrain, and assistive technology. For the purpose of brevity, assistive technology, specifically seating and wheelchairs, and the physical environment and their contribution to pain in adults with CP will be briefly addressed.

Many adults with CP use wheelchairs either part or full time; usage over time increases with the loss of locomotion skills noted in this population.9,11,16,26 Proper fitting of both manual and powered wheelchairs is important to the comfort and usefulness of such equipment. Persons with spasticity, scoliosis, and lower limb deformities often require specialized seating to insure support and comfort in the chair.54,55 Physical and occupational therapists provide recommendations and training related to seating and wheelchairs; assistive technology suppliers or vendors and rehabilitation engineers are involved in the process as well. The Center for Medicare and Medicaid Services now requires advanced certification from the Rehabilitation Engineering Society of North America for all professionals involved in provision of powered mobility equipment after April of 2008.56 These requirements were facilitated primarily by cost control concerns; however, there is evidence that wheelchairs cause pain in those who use them, underscoring the need for skilled prescription and fitting of such devices.57,58 Young adults in Evans et al.57 reported problems with jarring and insufficient support in their chairs. Not all pain from wheelchairs resulted from the fit of the system; aspects of the physical environment were cited as sources of pain, including potholes and poorly maintained sidewalks, uneven terrain, and difficulty fitting under tables and through doors. Accidents in wheelchairs comprised of tipping and collisions with furniture and people, have been reported by young adults and adult users.59 Increasing rates of obesity in CP increase challenges in seating comfort and limit the ability of caregivers to adjust the position of wheelchair users who cannot move themselves.40,41

Summary

The preceding overview has outlined some of the health parameters involved in adults with CP. Factors not reviewed that may also contribute to pain include aspects of living environments – availability and funding for personal care attendants, frequency of position change, access to appropriate seating systems, home accessibility and modification, worksite arrangement, and ergonomic variables.

Influence of pain on activity and participation

Many adults with CP experience pain, but does this pain influence daily activities, and, if so, how? Some variability in study results is evident. Participants in the Netherlands did not experience functional or social limitations from pain.18 Schwartz et al.14 noted that 49% of their sample reported that pain did not interfere with their activities or social participation. The follow-up investigation by Engel et al.12 revealed similar results, with only 14% of the sample reporting that pain interfered with social life and work. Jensen et al.60 found a slightly higher extent of pain interference, with 33% of the sample noting a moderate-to-severe impact of pain on daily activity. Significant associations between physical role function, low life satisfaction, and deterioration of functional skills and pain were reported in two other studies.13,26 Pain with ambulation in adults who walked independently and with aids was reported, which could limit community participation.19 In other studies, adults with CP reported having stopped walking due to pain, contractures, decreased strength and balance, and increased spasticity.16,26

Limited evidence suggests that while a large percentage of the adult population with CP does live with pain, its impact on their ability to perform daily activities is limited. Several authors have noted that only a small percentage of adults seek medical assistance with their pain,12,23 which may indicate that pain interferes with activities to only a limited extent. This may be due to the fact that coping strategies such as attention diversion and persisting in tasks, as described by Jensen et al.,60 or use of various pain treatment interventions are in fact effective for some persons.61 It is also possible that activity and participation are already limited by motor and cognitive impairments, especially for those with severe involvement. For these individuals, the impact of pain may be negligible in light of other existing restrictions. Given the diversity found in the CP population and the limited samples studied to date, there is still much work to be done to understand which individuals with CP are most affected by pain and the degree to which pain interferes with their lives.

Interventions for chronic pain

Studies regarding interventions to address pain in adults with CP are limited; medication solutions have the most evidence to support their use. This section discusses measures that adults with CP report using to manage pain, medical interventions, and other approaches to pain relief.

Engel et al.61 asked 64 adults with varied distributions of CP about how they managed chronic pain. More than 50% of the participants in this sample used non-steroidal anti-inflammatory medications to treat chronic pain; only one-third of participants used anti-spasticity medications or narcotic analgesics. All these medications reportedly had limited success. Healthcare approaches (counseling, biofeedback, relaxation training, cognitive therapy) and health-promotion approaches (exercise) were reported, but they were used less commonly than medication. Interestingly, the persons who used these methods were better satisfied with them than were individuals who used medications.

Medication interventions

Medication interventions for pain in CP include antispasmodics, antireflux agents, antacids, laxatives, and antidepressants.39 Because they are so commonly used, antispasmodic medications will not be addressed here.

Oral medications used in CP to address spasticity include the benzodiazepines, baclofen, tinzanidine, and dantrolene sodium.62 All produce sedative side effects that may or may not decrease with time. Associated problems with poor head and postural control as well as increased drooling are often seen. Nighttime use of the benzodiazepines has been shown to minimize drowsy side effects and decrease spasticity.63 No studies indicate that oral medications alleviate pain in adults.

Botulinum toxin A (BoNT-A) is frequently used to manage spasticity interfering with motor control and function in children and adults with CP. It is injected directly into selected muscles and acts by preventing the release of acetylcholine at the neuromuscular junction. Effects are evident in 1 to 7 days and last 2 to 6 months. Common side effects are injection-site discomfort, rash, and fatigue. The main purpose of BoNT-A is to reduce spasticity; in doing so, it may reduce pain associated with spasticity. In a randomized, double-blind, placebo-controlled study, Yelnik et al.64 demonstrated a significant decrease in shoulder pain after injecting BoNT-A directly into the subscapular musculature of adults after stroke. The best-described use of BoNT-A specifically for pain was found in its application to chronic neuropathic pain.65 In this randomized, double-blind, placebo-controlled trial, patients receiving BoNT-A demonstrated significantly decreased pain over time on a visual analogue scale up to 14 weeks after injection. Similar studies have not been carried out in individuals with CP; however, BoNT-A is a commonly used antispasmodic in the CP population, with anecdotal reports of pain reduction with its use.

Oral baclofen requires relatively high dosages since it does not cross the blood-brain barrier efficiently, and has side effects of drowsiness and muscle weakness. Baclofen delivered intrathecally (ITB) has been demonstrated to be effective in the management of spasticity, although concerns with adverse events have been noted.66,67 The effectiveness of ITB for pain management has been demonstrated.68 Slonimski et al.69 provide an extensive discussion of the uses of ITB for pain management across various health conditions.

Biofeedback

Using a multiple baseline design, Engel et al.70 studied the efficacy of biofeedback-assisted relaxation training to treat chronic pain in three adults with CP. Two participants had spastic diplegia and one had spastic quadriplegia. The intervention was received for six to seven 1-hour sessions with recommended home practice a minimum of twice daily. Pain documentation via participant log was carried out for 60 days after the intervention ended. Two of the three participants reported decreased pain intensity after the intervention, although electromyography documented no change in muscle activity.

Exercise

Exercise has been shown to reduce pain in individuals with a variety of health conditions. The evidence is most compelling in arthritis studies, where multiple randomized, controlled trials have demonstrated the positive impact of exercise on pain levels. Other conditions where exercise has been successfully used to manage pain include fibromyalgia71 and multiple sclerosis.72 Adults with mobility impairments who demonstrated higher levels of physical activity reported greater levels of community participation and better overall health than did participants who were inactive or had low levels of activity.73

Twenty-one adults with CP acted as their own controls in a three-phase study using an exercise intervention for pain management three times a week for 23 weeks.74 Eleven participants were ranked at GMFCS levels I to III, and 10 were ranked at GMFCS levels IV to V. Mean age was 39 years, with a range of 23 to 56 years. Outcome measures included pain intensity and location using a numerical rating scale and the Pediatric Quality of Life Inventory Multidimensional Fatigue Scale administered monthly. The intervention was delivered in a group setting and consisted of resistive and cardiovascular activities. Pain intensity was not altered to a statistically significant degree. Preliminary data analyses indicated the number of pain sites decreased significantly between the first two phases but this difference was lost at the end of the third phase. A significant interaction was seen between group (ambulatory/non-ambulatory) and disability from pain. Ambulatory adults with CP experienced less pain disability with exercise while non-ambulatory adults appeared to experience greater pain disability. Comparisons across body locations were made, with significant decreases in pain found in the head/neck/trunk and in the hips and knees, but not in the feet, ankles, or upper limbs. Significant improvements in all three fatigue subscales were noted as well. There were no differences by sex or GMFCS level.

In an earlier study that used a twice-weekly swimming intervention over 8 weeks, six adults with CP who lived in group homes demonstrated significantly decreased pain.75 All participants had motor impairments at GMFCS levels IV or V. While time in the pool was believed to be beneficial, another factor contributing to pain decrease suggested by this study was change of position from prolonged sitting. The participants typically stayed in their wheelchairs 12 to 14 hours.

Exercise options for persons with CP and other disabilities have been promoted for many reasons, including their effectiveness in managing obesity, their ability to improve strength and cardiovascular fitness, and their positive role in the prevention of secondary conditions.76–78 Adults with CP have reported exercise is helpful in pain management.15,20 Adults with GMFCS levels IV and V have successfully participated in exercise interventions.74 Although assistance was required for participation, these adults were able to use adapted devices and one-on-one assistance for resistive and cardiovascular exercise. The resistive and cardiovascular exercise options were easier to manage than the swimming intervention since clothing change and transfers were not required.

Challenges do exist at the community level for participation in exercise options. Not all fitness centers are accessible.78 Adults with severe impairments need assistance not usually provided by facilities, and uni-sex changing rooms are rarely available. Transportation is a barrier for adults who do not drive and do not have a family member who can take them to the exercise facility. Costs of membership, transportation, and personal assistance are other barriers.78 Finally, not all adults with CP are interested in exercise options. In spite of these challenges, preliminary studies suggest that exercise is an effective tool to manage pain.

Conclusions and recommendations

Evidence regarding the incidence and prevalence of pain in children and adults with CP is emerging and has been relatively consistent across studies. Adults with CP, regardless of level of impairment, do experience considerable pain, often from multiple sites. Little is known about pain in adults with CP and cognitive impairments. The degree to which pain interferes with life activities appears greater in a small subset of individuals, but why only some people experience such pain needs further study.

Potential sources of pain are numerous; however, careful study of the contribution of such sources is limited. Given the numerous pain sites reported across studies, it is likely that multiple factors contribute to individual pain levels. Sources of pain in adults with relatively less physical impairment in particular are not well understood. While health-related sources of pain have been hypothesized, and in some cases studied, there is a dearth of information on environmental contributions to pain. Examples of such contributing factors are infrequent position change due to lack of care providers, prescription of inappropriate equipment, prolonged waits for transportation and health services, poor-quality care in residential facilities, and the poor health status of aging parents who care for their adult children with CP.

Interventions for pain relief have been identified by persons with CP; however, these are not consistently used. Moreover, adults with CP do not appear to discuss pain management with their physicians. Medical interventions such as BoNT-A and ITB have support for efficacy in pain management. Other interventions, such as exercise and biofeedback, have limited but positive support for effectiveness.

There is an extensive need for both descriptive and experimental studies. Multi-site studies with large samples using objective measures, as well as self-report, are needed in order to better understand pain causality across the heterogeneous CP population. Interventions need to be developed that can be applied with limited, if any, assistance and at no cost so that adults with severe motor impairments who are living on restricted incomes do not need to depend on others for pain relief. Options such as meditation, relaxation, and guided imagery need exploration. Some aspects of pain may not need healthcare solutions but rather community planning and community services. Pain incidence, sources, and management are topics that clearly need extensive study so that today’s children with CP have a better quality of life in their adulthood.

Ancillary