Pain, 46 (1991) 247-264

247

0 1991 Elsevier Science Publishers B.V. 03~-3959/91/$03.50 ADONIS

030439~991001807

PAIN 01927

Clinical Section Review Article

The epidemiology of pain in children and adolescents: a review J.E. Goodman 0Department

a and P.J. McGrath b

of Psychology af Queen’s University, Kingston, Ont. (Canada), and ’ Deparfments of Psychology, Paediatrics and Psychiarry at Dalhousie UniLiersity,Halifax, NS B3H 4.U (Canada)

(Received 24 January 1991, revision received 25 March 1991, accepted 16 May 1991)

Introduction Epidemiology is the branch of health science that focuses on investigations of the distribution and determinants of disease or illness in defined populations. Epidemiological studies provide data crucial to the understanding of the etiology, natural history and impact, as well as the aggregation and transmission of a disease or condition. Such studies also document the interrelationships among risk factors associated with the syndrome being investigated. Understanding disorders in these terms is a prerequisite to their prevention and may be invaluable to their treatment. In chronic pain, as in most other chronic disorders, a major aspect of the burden of pain is the resultant changes in lifestyle that often accompany pain. The World Health Organization [65] has developed a model for the impact of diseases that has recently been applied to pain [36]. According to this model, four planes of experience can be described. The first plane is the underlying disease or disorder. The second plane, termed impairment, refers to symptoms including pain. The third plane is disability, which describes those activities that are curtailed as a result of the problem. The fourth plane, handicap, is the restriction of social roles. The relationship between planes of experience is not linear and the different factors influencing pain, disability and handicap are not well understood. Because pain, regardless of etiology, may be associated with extensive disability and handicap, the epidemiology of pain, disability and handicap merits considerable attention. Epidemiological studies of disability

Correspondence foe: P.J. McGrath, Departments of Psychology, Paediatrics and Psychiatry, Dalhousie University, Halifax, NS B3H 4J1, Canada.

and handicap due to pain may assist in finding clues to the origin of the burden of pain. From our currently limited understanding, it is clear that pain itself is only part of the cause of handicap accompanying pain. Little is known about the prevalence of pain and resulting disability and handicap in a general adult population [lo]. Data pertaining to the risk factors associated with pain and associated disability and handicap are also weak and speculative. This is also true for pain in children and adolescents. Pain is a health care problem which carries severe personal and economic consequences. The Nupria Pain Report [22], conducted in the United States, estimated that in 1985 a half a billion work days were lost directly because of pain among those employed full time. The physical and psychological suffering endured by those experiencing pain is, of course, more difficult to measure. Pain in children and adolescents does not create the same economic hardships as does adult pain 1251. Motivation for a more comprehensive understanding of pain in childhood and adolescence is, in this respect, lacking. However, a better understanding of pain in children and adolescents is warranted because of the burden of suffering accompanying pain and because such an understanding might provide further clues to the origins of adult pain problems. Clinicians have suggested that many types of pain problems aggregate or cluster in families. For instance, Apley and Hale [l] referred to the concept of the ‘painful family’ in their study of recurrent abdominal pain in children. The transmission of pain, disability and handicap from one family member to another could occur through biological or psychological mechanisms. However, little substantive data exist which explore the mechanisms by which pain, disability and handicap are transmitted within families and across generations [36]. Epidemiological studies of pain in

24X

families could determine if aggregation of pain problems occurs and might identify factors contributing to this phenomenon. Most epidemiological studies investigating pediatric pain have addressed specific pain conditions (e.g., pain from headache, recurrent abdominal pain) rather than providing a comprehensive analysis of pain problems in childhood and adolescence. Moreover, because of limitations in their design and methodology, as well as differences in the defining criteria for pain, the data they provide are often conflicting. For instance, Deubner [ 131 reported the prevalence of headache in adolescent females was 82%, whereas in a similar study Passchier and Orlebeke [45], reported the prevalence as approximately 11% for the same age group, even though both studies used the same questionnaire used by Waters [61]. The differences in the reported prevalence rates can be attributed to differences in the definition of headache used and the time frame about which children were questioned. This review will focus on the findings and the methodological issues pertaining to epidemiological studies of pediatric pain. Three types of methodological problems are particularly important: issues regarding the reliability and validity of pain measures, issues regarding the breadth of measurement, and sampling issues. These are summarized in Table I. By definition, pain is a subjective phenomenon [371 and, thus, any measure that is not based on self-report will likely yield inaccurate results. Many studies of children’s pain have based estimates of the prevalence of pain on parental reports [e.g. 6,41,59,66]. Although there is some correspondence between children’s pain ratings and parents’ pain ratings [48], they are not identical. In addition, the use of pain measures that were designed for adults may not be appropriate. The construction and psychometric validation of developmentally appropriate methods of pain measurement has occurred only recently and, as a result, the measures used in many epidemiological studies are of questionable validity. Most epidemiological studies of pediatric pain have used simple occurrence or non-occurrence of pain. This approach has serious problems because a very minor pain occurring once a week may be regarded the

TABLE

same as a very severe pain happening all of the time. Obtaining measures of severity, frequency and duration can correct this problem. An alternative strategy is to define what constitutes a clinical case and to examine the prevalence and correlates of ‘caseness.’ All epidemiological studies of pain in children and adolescents have been retrospective. In other words, respondents are asked to recall their pain experienced within a certain time frame in the past. The length of this time frame is important. Extremely long time frames (e.g., 1 year or longer) are likely to be unreliable whereas too short a time frame (e.g., 1 day) may yield less meaningful data. An alternative strategy is to use a pain diary which requires prospective recording of pain 3 or 4 times a day [48]. Diaries have the disadvantage of requiring more time from respondents and thus may contribute to lower participation rates. The breadth of what is measured will impact on the usefulness of the study. For example, the propensity to measure only one pain at a time precludes an understanding of the inter-relationship between pain problems. In addition, measures of disability and handicap should be incorporated into pediatric pain studies. Unfortunately little attention has been devoted to the development and validation of measures of disability and handicap. Both McGrath’s [321 and Varni’s [60] pain assessment techniques contain measures of disability but neither has been validated. To date, no measures of handicap other than school absence have been developed. Furthermore, although cross-sectional studies are invaluable, longitudinal studies are needed to disentangle developmental changes from secular trends. Sample representativeness is another issue which must be critically appraised. Many researchers have used samples which are inappropriate for the epidemiological investigations they have planned [e.g., 58,641. There is little empirical advantage in employing samples of convenience or samples the characteristics of which are not known because generalization to populations is not valid. Many studies have examined subjects of a very narrow age range [e.g., 3,17,66], which does not allow for examination of developmental changes in pain problems. Because childhood and adolescence is a relatively rapid period of growth and development,

I

METHODOLOGIC

CONCERNS

WlTH

PEDIATRIC

I, Measurement of pain Self-report Age appropriate measures Known psychometric characteristics Measures of severity, frequency and duration Measure of caseness Time frame specified and reasonable

PAIN STUDIES 2. Broad based measurement Multiple pains measured Measurement of disability Measurement of handicap Measured across generations Longitudinal measures

3. Sample Described Representative Large enough Age spread sufficient

249

inclusion of a broad age range of subjects may be more useful than focusing on a single age group. Because of differences in the factors cited above, epidemiological studies are not easily compared. We believe that attention to methodological issues will help evaluate which data are accurate and valid and will also highlight what is needed in future studies. The epidemiological studies which have investigated specific pain problems in children and adolescents can be classified into 2 broad categories: those investigating pain in a non-clinical population (e.g., schoolchildren) and those investigating pain in clinical or special populations (e.g., outpatients of a neurology unit or competitive athletes). Determining which studies to include in the clinical or special population section is difficult because in many studies it is not clear whether the study in question meets the criteria for an epidemiological study. Non-clinical and clinical studies will be separately evaluated. To facilitate organization, studies will be reviewed according to type of pain disorder and will proceed historically.

Pain in non-clinical

settings

We will include in this section all studies in which the sample used is drawn from a non-clinical population, that is the major defining characteristic of the sample is not medically or clinically oriented. Pain is a symptom which prompts many, but not all, to seek medical attention [lo] and there may be substantial differences between those who seek medical attention and those who do not. In non-clinical studies, whether participating subjects sought medical attention for their pain is irrelevant, unless it is one of the dependent variables being measured. The studies investigating the incidence and prevalence of pain in children and adolescents in non-clinical settings are summarized in Table II. Head pain Headache

and/or migraine. Headache is the most extensively investigated pain problem of childhood and adolescence. Investigations into the prevalence of migraine in children and young adults were pioneered by Vahlquist [59] and by Hughes and Cooper [23]. Although both studies are interpretable and interesting, several limitations must be noted. Both studies neglected to specify the sampling frame or the time period about which subjects were queried. Moreover there was extensive reliance on reports from mothers as the sole source of information. Both studies had a reasonable sample size but age groups of particular interest were omitted. Vahlquist’s criteria have been adopted by many subsequent researchers [e.g., 651-531.

Vahlquist [59] found that the prevalence of migraine in those aged lo-12 years was 4.5% and in those aged 16-19 years was 7.4%. No sex differences were found. The age of onset for migraine was before the age of 6 years for almost half of those aged lo-12 years. Hughes and Cooper [23] determined the prevalence of headache in 5-, lo- and 14-year-old males and females was 3%, 17.5% and 12.3%, respectively. In 16-year-old females the prevalence was 14.2%. They attempted to establish criteria for ‘caseness’ based on severity and handicap. Mild headache, defined as occurring ‘at odd times,’ accounted for almost ten times the number of headache complaints as severe headache, which was defined as painful enough to ‘reduce the child to tears or to bed.’ Bjlle [6] reported the first findings of an ambitious longitudinal study on the prevalence of migraine begun in 1955. Questioning the parents of 9059 children, aged between 7 and 15 years, the prevalence of migraine in the age groups 7-9, 10-12, and 13-15 years was 2.5%, 4.6% and 5.3%, respectively. In the same study, a special group of 73 children, aged 9-15 years with more pronounced migraine, and a control group of 73 children of the same age without headache symptoms were followed for 23 years. Bille [7] reported that 60% of those in the pronounced migraine group were still experiencing migraine attacks 23 years later but, in half of these cases, the attacks were neither as frequent nor as severe as in childhood. In comparison, only 11% of the control group had begun to suffer from migraine during this period. In subsequent follow up [Sl, the original group of children, most of whom were now in their thirties, were questioned about migraine symptoms in their children. Of the 90 children born to the original pronounced migraine group, 15 suffered from migraine. Although Bille’s work is important, especially in understanding the natural history of pronounced migraine, self-report data were not used for the original large sample and the smaller sub-sample. Another, more simple investigation of migraine found a slightly higher prevalence than did Vahlquist [59] and Bille [6]. Dalsgaard-Nielsen [l 11 directly interviewed 2027 children and adolescents aged 7-18 years and their parents on several occasions and found ‘genuine’ migraine present in 73 boys and 71 girls, or 7.1% of the total sample. Accepted diagnostic criteria for migraine prescribed by the Ad Hoc Committee on the Classification of Headache were used. A second group of 390 adult migraine sufferers (138 males, 252 females) were included in the analysis and the age of onset of migraine symptoms in both groups was lower in males than in females. Waters [61] assessed the prevalence of headache in several Welsh communities using a sample of adolescents and adults. For the purpose of our review, only results pertaining to the youngest age group will be

250 TABLE

II

SURVEYS

EXAMINING

Reference

THE EPIDEMIOLOGY

OF PEDIATRIC

Disorder

Sample

migraine

I. n = 1236, 10-12 years II. n = 1373, 16-19 years Ill. n = 792, 20-29 years

PAIN

IN NON-CLINICAL

Sample

POPULATIONS

representativeness

Critical

summary

1. Heud pain Vahlquist

Hughes

[59]

and Cooper

[23]

Bille [6]

Dalsgaard-Nielsen

et al.

aged

Sampling frame was not detailed; sex distribution was not given

Questioned mothers of youngest subject group; attempted to identify accompanying symptoms and potential triggers of migraine: prevalence in each of the 3 subject groups was 4.5%, 7.4% and 5.90/c, respectively, hut time frame for recall was unspecified.

aged aged

headache

1. n = 1721, aged 5 years Il. n = 969, aged IO years III. n = 1002, aged 14 years IV. n = 119, aged I6 years

Sampling frame was not detailed; sex distribution was not given

Poorly designed and reported study, attempted to identify symptoms accompanying headache and locus of pain, fails to use self-reported data; prevalence in each of 4 subject groups was 3%. I7.5%, 12.3% and 14.2%. respectively, but lime frame for recall was unspecified.

migraine

n = 9059, aged 7-15 years

Inclusion criteria for sample was carefully detailed; used entire population of children aged 7-15 years in city where survey was conducted

Failed to use self-reported data; prevalence of migraine in those aged 7-9, lo-12 and 13-15 years was 2.5%. 4.6% and 5.3(X, respectively, hut time frame for recall was unspecified.

migraine

1. n = 2027 (1075 males, 952 females) aged 7-18 years Il. n = 390 (I38 males, 252 females) aged 18-69, migraine patients

Sampling

frame

was well detailed

Children and parents were directly interviewed on several different occasions; assessed prevalence in relation to age of onset, sex distribution; overall prevalence was 7.1%. hut the time frame for recall was unspecified.

[Ill

Waters

[61]

headache

n = 963, aged years

15-75

Sampling

frame was not detailed.

Respondents gave I year retrospective account of their headaches; reported prevalence was 71% (ml and 92% (fl, for the age group of 15-34 years.

Waters

[62]

migraine

n = 1718, age range not reported

Sampling

frame was not detailed

Accompanying symptoms assessed in the study were not defined, control group used was not a control but similar group with headache rather than migraine: reported prevalence was 4.9% (ml and 9.1% (ff.

Sillanpaa

[51]

migraine/ headache

n = 4235 (2195 males, 2039 females). aged 7 years

Used restricted age range of subjects although method of sampling was carefully documented

Questioned both children and mothers to assess children’s headache/migraine; prevalence of headache in past month was 11.5%.

Deubner

1131

migraine/ headache

originally targetted 780 subjects; 600 remained at time of follow -up (297 males, 303 females), aged lo-20 years

Sampling frame was not well detailed, lost 23% of sample at time of follow-up

Questioned both target children and parents, sound in methodology and statistical analysis; in past year, prevalence of migraine was 15.5% (ml and 22.1% (ff and of headache was 74% (ml and 82% (fl.

migraine

n = 12,543 males and n = 3242 females. aged 10-18 years

Sampling tailed

Directly questioned subjects, used accepted migraine/ headache criteria; prevalence was 3.3’1: (ml and 2.5% (f).

Sparks

[54]

frame

was not well de-

251 TABLE II (continued) Reference

Disorder

Sample

Sample representativeness

Critical summary

Bille 171

migraine

n = 73 (32 males, 41 females), aged 9-15 years in 1955, plus equal number of matched controls

See Bille 161

Continuing longitudinal study of migraine begun in 1955, does not use self reported data; for reported prevalence, see Bille [6].

Bille [8]

migraine

n = 73 (32 males, 41 females), aged 9-15 years in 1955, plus equal number of matched controls

See Bille [6]

Similar methodological problems as Bille [7]; for reported prevalence see Bille [61.

Sillanpaa [52]

headache

n = 3784 (1911 males, 1873 females), aged 13 years

Sampling frame was well detailed but used a very restricted age range of subjects

Defined all important variables, used self-reported data, assessed prevalence and frequency of headache/migraine: overall prevalence for past year was 12%.

Sillanpaa [53]

migraine

n = 2921 (1473 1448 males, aged females), 14 years who participated in Sillanpaa 1511

See Sillanpaa 1511

Seven-year longitudinal follow-up on frequency and prevalence of headache and migraine; no major methodological flaws; prevalence for those aged 7 and 14 years was 2.7% and 10.6%, respectively.

Collin et al. [9]

headache

I. n = 686, aged 5-14 years II. n = 1451, aged 5-19 years

Reported exclusion criteria for sampling frame but does not give age or sex distribution of sample

Attempted to assess handicap of headache in terms of number of days absent from school, does not use self-reported data, participation rate was only 79%; time frame for recall by parents was 12 weeks; prevalence of school absence was 3.7%.

Passchier and Orlebeke 1451

headache

n = 2286 (1132 1049 females, males), aged lo-17 years

Used a random, stratified sampling technique to obtain a 4% sample of all children aged lo-17 years in city

Used self-reported data to assess prevalence and frequency of headache; time frame for recall was 4 weeks: prevalence ranged from 9-12s in males, and was 11% for females, depending on age group.

Linet et al. [30]

headache

n = 10,169 (4394 males, 5055 females) aged 12-29 years

Sampling frame was well detailed

Directly questioned subjects using an interview, follow-up of same occurred 4-8 weeks later: for those aged 12-17 years, prevalence of migraine was 3.8% (ml and 6.6% (f) and for headache was 55.9% (ml and 73.6% (fl; time frame for recall was 4 weeks.

Wigdorowicz-Makowerowa et al. [64]

oral pain

I. n = 2100, lo-15 years

aged

Sampling frame was not detailed, no information about sex distribution of respondents

Did not examine pain, per se, subjects in the lo-15 year age group were 1 of 5 subject pools examined, did not use self-reported data, interview process not detailed; reported prevalence (painful bruxism) was 2.6% at time of study.

Nilner and Lassing I421

jaw/oral

n = 440 (218 males. 222 females), aged 7-14 years

Sampling frame was well documented

Examined the signs that indicate disease rather than pain, per se: as an accompanying symptom, prevalence of headache was 19%.

pain

252 TABLE

II (continued) Sample

Sample

Critical

representativeness

summary

Reference

Disorder

Egermark-Eriksson

jaw/oral

pain

n = 402 males, 1 I, (70 males), males, years

aged 7, (62 74 females), males, 61 fe15, (76 5Y females)

Sampling

frame was not detailed

assessed present Primarily mandibular dysfunction, looked at pain only as an accompanying symptom; 33% reported muscle tenderness, 2% reported pain on movement of jaw.

jaw/oral

pain

n = 1342 (674 males 668 females), aged 6-17 years

Sampling

frame was not detailed

Assessed present TMJ (temiporomandibular joint) pain and muscle tenderness as signs and symptom of TMJ dysfunction: prevalence of TMJ pain was 7%; 17% reported muscle tenderness.

n = 1000 (528 males,

Sampling frame was not detailed. did not report age range of subjects

Carefully defined syndrome; reported age of onset, age distribution, and accompanying symptoms, questioned both target children and mothers; time frame for recall was 1 year; prevalence was Y.SV (ml and 12.3% tf). Attempted to assess handicap suffered by child. parental management of pain, sociocultural differences but fails to use self-reported data; time frame for recall was unspecified; overall prevalence was 10.1%.

et al. [14]

Riolo et al. [49]

2. Stomach /abdominal

pain

recurrent abdominal pain

Apley and Naish [2]

472

females),

aged

up to 15 years

Faull and Nicol [17]

recurrent abdominal pain

n = 145 (72 males, 73 females) aged 5-h years

Used very restricted subjects

Golub

dysmenorrhea

n = 16,183 attending school

females high

Age range not reported

dysmenorrhea

1. n = 5458 aged 10-20 II. n = 331 aged 13-20

females, years females. years

dysmenorrhea

dysmenorrhea

et al. [21]

Widholm

[63]

Klein and Litt [27]

Teperi and Rimpela

[57]

age range

of subjects

Reported age distribution ple but failed to provide mographic information

of

surveyed

Failed to define syndrome or additional dependent variable used in the study; time frame for recall was unspecified; prevalence of occasional and frequent pain was 37.4% and 30.2%, respectively.

of samany de-

Reported age of onset of symptoms; measured handicap in terms of number of days absent from school: time frame for recall was unspecified; prevalence of occasional and invariable pain was 3X.5% and 13.2%.

n = 2699 females, aged 12-17 years

Used data collected by Natl. Center for Health Statistics 1966 1970, reported that sample consists of cross-sectional natl. probability sample

Questioned target subjects as well as parents. also used data ohtained from physical examinations, psychological testing, interviews with school personnel; time frame for recall was unspecified: overall prevalence was 59.7%.

n = 3370 females, aged 12-18 years

Sample was derived from the Natl. Population Registry in Finland. sampling frame was well documented

U\ed self-reported data to assess prevalence of pain and associated medication use and school absenteeism; questionnaire used was previously validated; time frame for recall was 6 months; prevalence.

70%.

by age.

ranged

from

4X-

253 TABLE

II (continued)

Reference

Disorder

Sample

Sample

Naish and Apley [41]

non-arthritic limb pain

1. n = 721 non-clinical subjects II. n = 54 clinical subjects (age ranges in I and 11 not reported)

Only studied those reporting limb pain, failed to use appropriate control group

Failed to use self-reported data: method of pain measurement was not documented; time frame for recall was unspecified; reported incidence was 4.2%.

0ster

‘growing’

n = 2178 (1062 males, 1116 females), aged 6-19 years

Sampling

Directly questioned target children but interview and assessment were not validated; reported age range of 6-19 years but no information given on those subjects older than 17 years; time frame for recall was unspecified; prevalence was 12.5% (m) and 18.4% (f).

I. n = 446 (227 males, 219 females) non-clinical subjects, aged 13-17 years. II. n = 52 (14 males, 38 females) clinical subjects with knee pain, aged 13-36 years

Non-clinical group was randomly inclusion criteria for sampled, clinical group were well detailed but age ranges of the 2 groups differed markedly

Used both self-reported data and physical examination; attempted to assess knee pain in relation to physical activity level; time frame for recall was 1 year: overall prevalence was 31%.

n = 446 (227 males, 219 females) aged 13-17 years

Sampling

Used self-reported data; assessed back pain with respect to age of onset, whether medical attention was sought, related school absences; time frame for recall was unspecified; overall prevalence was 26%.

low back pain

n = 1715 (840 males, 875 females) aged 7-17 years

Sampling frame was well detailed and designed specifically to facilitate follow-up

Prospective study; questionnaire used was previously validated against accepted criteria; reported prevalence was 33%.

recurrent dominal headache, pain

abpain, limb

n = 2178 (1062 1116 females, males), aged 6-19 years

Only 635 of the original 2178 subjects were followed for 5 years, not reported how this group was chosen out of the original sample

Age range reported at 6-19 years but no data given for those older than 17 years; defining criteria for different types of pain was not reported; time frame for recall was unspecified; prevalences of headache, abdominal pain and growing pain were 20.6%, 12.3% and 15.5%. respectively.

stomach headache

ache,

n = 308, aged 3 years

Used very restricted age range of subjects, sampling technique not documented

Failed to use self-reported data: method of pain assessment of children by mother not reported, attempted to relate symptoms in child to ‘psychosocial stressors’ affecting mother; time frame for reported recall was 4 weeks: prevalences of stomach ache and headache were 9% and 3%. respectively.

representativeness

Critical

summary

3. Limb pain

and Nielsen

Fairbank

[44]

et al. [15]

pains

knee pain

frame was well detailed

4. Back pain Fairbank

Balagui

et al. [16]

et al. [4]

adolescent pain

back

frame was well detailed

5. Multiple pains

Bster

1431

Zuckerman

et al. [66]

254 TABLE

II (continued)

Reference

Disorder

Sample

Sample

Aro et al. [3]

psychosomatic symptoms

(1002 n = 2246 males, 999 females), aged 14- 1.5 years

Used very restricted age range of subjects, but included in the sample approximately 98% of all children of that age in the city

discussed. The questionnaire used, although not reportedly validated at the time of the study, has since been used in many subsequent studies of the epidemiology of headache [e.g., 13,451. Each respondent provided a retrospective account of their headache episodes occurring in the past year. For respondents aged from 15 to 34 years, 71% of males and 92% of females reported at least one headache episode. In a similar study, Waters [62] studied the prevalence of migraine in childhood and its relationship to accompanying symptoms including bilious attacks, travel sickness and eczema. Accepted diagnostic criteria were used to define an episode of migraine, but none defining the accompanying symptoms were reported. The prevalence of migraine in males (4.9%) was almost twice as high in females (9.1%). The accompanying symptoms assessed were more common in migraine sufferers than in those with non-migrainous headache or no headache. The age range of the respondents and the sampling frame were not reported. Using a carefully designed research strategy, Sillanpaa [51] reported the initial findings of an ambitious longitudinal study on headache and migraine. Using both a physical examination and an interview with the child and a parent, headache symptoms were assessed in a group of 4235 7-year-old children. The prevalence of headache was 37.7% and of migraine 3.2%. Sillanpaa [52] assessed headache in 3784 13-year-old adolescents in a similar manner. A total of 82% of the subject group had experienced some form of headache in the past year. The prevalence of migraine in females (15.1%) was almost twice as high as in males (8.1%). By comparing these results to Sillanpaa [51], changes in headache and migraine with age were assessed. In Sillanpaa [53], a follow up of 2921 children from the original sample [51], now 14 years of age, were again questioned regarding headache and migraine symptoms. The prevalence of headache and migraine episodes was found to have increased to 69% and lO.h%, respectively. Whereas in the 7-year-old group migraine was more common in boys; in the 14-year-old group, migraine was present in almost twice as many

representativeness

Critical

summary

Did not examine pain, per se, but some of the psychosomatic symptoms assessed are painful; e.g.. prevalences of headache, abdominal pain and heartburn were 7.2% (m) and 14.9% tf), 2.5% (m and f). and 1.2% Cm) and 0.7% (0, re\pectively.

girls (8.6%) as boys (4.7%). Accepted diagnostic criteria were used. Deubner [ 131 investigated various accompanying symptoms and risk factors associated with headache and migraine, including age, social class, menstrual status, cyclical vomiting, travel sickness and recurrent abdominal pain. Under the interviewer’s supervision, a sample of 600 subjects (297 males, 303 females; aged lo-20 years) completed the Waters [61] questionnaire which was modified for use with children. The prevalence of migraine was 15.5% in males and 22.1% in females. At least one episode of headache was found to have occurred in the previous year in 74% of males and 82% of females. There was no significant relationship between presence of migraine or headache to menarcheal status, social class, travel sickness, abdominal pain or cyclic or nervous vomiting. In Sparks [54], 12,543 male and 3242 female students aged lo-18 years under the jurisdiction of the Medical Officers of Schools Association were studied. The method by which migraine sufferers were identified was inappropriate; bulletins were posted in various schools of 500 pupils or more, asking those that had experienced migraine to complete a questionnaire. There was no information available on the validity of the questionnaire. The prevalence of migraine in males was reported at 3.3% and in females at 2.5%. The low prevalence of migraine can be attributed to the way migraine sufferers were selected for participation. Collin et al. [9] present the only data on school absence due to headache. School absence may be the simplest ‘hard’ measurement of handicap in children. Subjects from two carefully defined geographical areas in England (686 from Faringdon and 1451 from Abingdon; aged 5-19 years) were chosen to participate in the study. In the Faringdon study, class registers were inspected for two 12-week periods and the parents of each absent child were mailed a questionnaire enquiring about the reason for each noted absence. Seventynine percent of the questionnaires were returned and 3.8% of the students missed school because of headache accounting for approximately 1% of total school ab-

255

sences during this time. Most of the absences (85%) lasted no longer than 1 school day and only 15% of absences were for periods of longer than 1 school day. In the Abingdon study, the number of children attending the school sick bay were recorded for one 12-week period. These were divided into 2 categories: those treated and sent back to class and those sent home. Of the 265 children that were treated and sent back to class, 20% (17 males, 36 females) had complained of headache. Of the remaining 69 children who were sent home, 10% (4 males, 3 females) were sent home because of headache. In the 1Zweek period, only 0.5% of all children participating in the Abingdon study were sent home because of headache. Passchier and Orlebeke [45] used random stratified sampling to select a 4% sample from all children in Amsterdam, aged lo-17 years, to participate in their study of headache and its relationship to stress. Using a modified and validated version of Waters’ Headache Questionnaire [61], 1132 males and 1049 females were asked about the nature and frequency of occurrence of their headaches. Approximately 10.5% of males and 11% of females reported headache at least once every month. Stress was the most frequently reported precipitating factor of headache, with 30% of elementary students and 40% of secondary students reporting it as the primary cause for their headaches. Linet et al. [30] conducted a population-based telephone interview survey of headache and migraine among 10,169 subjects, aged 12-29 years. The questionnaire was carefully described, and a response rate of 94.2% was reported but no information on its psychometric properties was provided. A smaller, unspecified number were questioned about medication use in a follow-up 4-8 weeks later. In the 12-17 year age group, 3.8% of males and 6.6% of females had experienced migraine in the previous 4 weeks. The prevalence of headache of any type in the same period of time was 55.9% in males and 73.6% in females. Headache is the most carefully studied pain in children. Not only are there excellent estimates of prevalence [45,51-531, but we also have an initial understanding of the correlates of headache [45,61,62] and the extent of handicap due to headache [9]. Studies examining the specific activities interfered with by headache and good co-morbidity studies are lacking. Oral/dental

pain

Oral pain includes pain experienced in the areas surrounding the mandible, the temporomandibular joint, the teeth, the mouth and the lips. All studies of oral and dental pain lack standardized diagnostic criteria and rigorous methods of measurement. In addition, samples are often not representative or even described [ 14,49,64]. In a poorly designed study, Wigdorowicz-Mako-

werowa et al. [64] examined functional disturbances of the masticatory system in 5 different subject groups, one of which was 2100 school children aged lo-15 years. None of the dependent variables were adequately defined. Pain, per se, was not measured but a diagnostic interview was used to assess the point prevalence of ‘painful bruxism’ which was found in 2.6% of the children examined. In a more carefully designed study, Nilner and Lassing [42] examined the prevalence of functional disturbances and diseases of the stomatognathic system. From school registers in Malmii, Sweden, 440 children (222 males, 218 females; aged 7-14 years) were randomly selected for participation in the study. Subjects were interviewed about symptoms of masticatory system dysfunction and then clinically examined in 1 of 2 school dental clinics. Nineteen percent of respondents suffered from recurrent headache (defined as more than 1 headache episode/week), 64% reported pain on palpation of the temporomandibular joint muscles and 39% reported pain on palpation of the temporomandibular joint itself. Egermark-Eriksson et al. [14] studied the prevalence of bruxism and of painful signs and symptoms of mandibular dysfunction in a group of 402 children aged 7, 11 and 15 years (208 males, 194 females). The examination process was carefully documented. Pain was assessed only as a symptom accompanying mandibular dysfunction. One third of respondents reported muscle tenderness and 2% claimed pain on movement of jaw, In an effort to acquire a better understanding of the origins of temporomandibular disorders, symptoms of temporomandibular joint dysfunction were investigated from a developmental perspective in a study conducted by Riolo et al. [49]. Pain was examined as a subjective symptom and clinical sign of dysfunction. In a crosssectional survey of 1342 subjects (674 males, 668 females; aged 6-19 years) 17% reported muscle tenderness upon palpation and 7% reported pain in the temporomandibular joint itself. The most important requirement of future work in this area is the development of valid standardized measures. Stomach and /or abdominal pain Recurrent abdominal pain. Researchers

and clinicians have generally agreed to use a ‘case’ definition of recurrent abdominal pain as pain with no known organic cause that recurs at least 3 times, lasts at least 3 months and interferes with activities [2]. Apley and Naish [2] conducted the first epidemiologic study by directly questioning (in the presence of their mothers) a random sample of 1000 school children (528 males, 472 females), the majority of whom were aged between 5 and 15 years of age. The prevalence of recurrent

abdominal pain occurring in the preceding year was 9.5% in males and 12.3% in females, but the prevalence changed quite dramatically according to age, with a peak prevalence of 28% at 9 years of age in females and of 14% at 6 years of age in males. Recurrent abdominal pain in 494 children aged 5-6 years was studied by Faull and Nicol [171. Based on a screening interview, the children were defined as being a ‘case’ (97 children) or not (342 children) as defined by Apley and Naish 121.A more intensive study of the disorder was conducted with a sub-sample of 145 children (72 females, 73 males) from the original sample of 494 children, comprised of all high risk and 48 low risk children. One or both parents of each participating child were questioned and the parents and teachers completed a questionnaire about abdominal pain. The prevalence of recurrent abdominal pain was 24.526.9% and pain appeared to be precipitated by worries about school, overeating or ‘trying to get out of something.’ Being a case was associated with externalizing behavior problems, social support of the mother, difficulties in settling in to school and living in a council flat. The prevalence of abdominal pain in this sample was very high. However, a 6-9 month follow-up of 26 of these cases [12] found that all but one were substantially pain free. The epidemiology of recurrent abdominal pain has received little attention but the two studies reported here, although flawed, are good examples of careful research. Dysmenorrhoea. Dysmenorrhoea is characterized by cramping pain in the lower abdomen at the time of menstruation, often accompanied by nausea, vomiting, edema and headaches 1271. Four studies [21,27,57,63] have examined the prevalence of dysmenorrhoea. The earliest study by Golub et al. 1211was seriously flawed by measurement problems but subsequent studies have been methodologically quite strong. Widholm [63] questioned 2 sets of adolescent females to assess the menstrual patterns of adolescents and the relationship between menstrual pain and school absence. The first series included 5485 females aged 10-20 years and the second series included 331 females aged 13-20 years. Of all respondents participating in the study, painful menstruation occurred in 7.2% during the first year following menarche and increased to 26% by the fifth year following menarche. In the second series of respondents studied, the prevalence of school absence due to dysmenorrhoea was 23.4%. Klein and Litt 1271 used a cross-sectional national probability sample of 2699 females aged 12-17 years obtained from the National Center for Health Statistics survey. Of the 59.7% of subjects who reported discomfort or pain in connection with menstruation, 14% described it as severe. School absence because of

dysmenorrhoea was reported by approximately t5% of subjects. Teperi and Rimpela [57] studied the prevalence of dysmenorrhoea and related school absence in a random sample of 3370 adolescent females aged 12-18 years. A validated questionnaire about the severity of menstrual pain, related medication use and related school absence was mailed to each participant. Eightyseven percent of those approached completed and returned an unspoiled questionnaire. Approximately 50% of lZyear-olds and over 80% of 18year-olds reported some degree of menstrual pain. Both the use of medication and rate of absenteeism in school correlated positively with severity of pain. The prevalence of dysmenorrhoea and associated school absence is well documented but the correlates of these phenomena are less well understood. Limb pain ‘Growing’ pain. Growing pain is a colloquial

term which is often used to describe pain occurring in the limbs during a period of skeletal growth in the absence of any other organic disease or physical injury. The causal relationship between growth and limb pains has never been established. In a study of non-arthritic limb pains, Naish and Apley [411 compared a group of 721 children sampled from a local school health clinic with a similar group of 54 children suffering from limb pains on a number of variables including family history and the time during which the pains were most often experienced (nocturnal vs. diurnal). In the first group, the researchers questioned both children and their mothers and found that 4.2% were experiencing limb pains. There was no significant difference between the first and second group in social class, frequency of respiratory infections, allergic disorders and a number of other variables. However, in the group with limb pains there was a greater number with a family history of rheumatic complaints. No demographic data were available for either of the groups studied and the sampling method was not adequately detailed. Bster and Nielsen 1441sampled 2178 children (1062 males, 1116 females; aged 6-19 years) who were undergoing routine examinations at a school health clinic in Copenhagen. Each child was directly questioned about ‘growing’ pains, defined as pains occurring in the arms or legs. Of all children sampled, 12.5% of males and 18.4% of females reported such limb pains, although the time frame for recall was unspecified by the researchers. The work on growing pains is poor and little is known about this condition. Knee pain. The prevalence and etiology of knee pain or pain in the patellofemoral joint was examined by Fairbank et al. [15] who measured joint mobility and

257

lower limb morphology of 446 school children (227 males, 219 females; aged 13-17 years). Each participating child also completed a questionnaire about activity level and knee pain experienced in the past year. In this group, the overall prevalence of knee pain was 31% and children were more likely to experience knee pain if they regularly participated in sporting activities. These results were compared to data from 52 hospital outpatients (14 males, 38 females; aged 13-36 years) suffering from anterior knee pain. In this group it was also thought that chronic overloading of the knee joint was the dominant contributing factor associated with knee pain. Back pain

Back pain occurring in childhood or adolescence has not been widely investigated. Fairbank et al. [16] studied back pain in 446 adolescents (227 males, 219 females; aged 13 to 17 years). Participants were physically examined, asked to identify the site of their back pain symptoms and completed a back pain questionnaire. A history of back pain was reported by 26% of the subjects and was most commonly reported by those who avoided sports and other physical activity. The time frame for recall of symptoms was unspecified. Balagut et al. [4] used a questionnaire, designed and validated according to accepted methods, with 1715 schoolchildren (840 males, 875 females; aged 7-17 years). The overall prevalence of low back pain was 33% with a sharply increased prevalence with age to 71.3% in 15year-olds. The time frame for reporting was the child’s lifetime. Very few children are seen clinically for back pain [26]. The high prevalence rates in these studies are misleading because there are no measures of frequency, severity and duration and there are no accepted ‘caseness’ criteria. These studies suffer as well from long or unspecified time frames and thus little can be said about the meaningfulness of back pain reported in these children. Multiple pain problems

Few studies have attempted to measure the prevalence of more than one type of pain simultaneously. In an 8 year longitudinal study, Oster [43] attempted to measure the prevalence of recurrent abdominal pain, headache and growing pains in a sample of 2178 children (1062 males, 1116 females; aged 6-19 years). Diagnostic criteria for the 3 disorders were not reported, many procedural variables were not described and the time frame for recall was not specified. The reported prevalence of recurrent headache, abdominal pain and growing pain was 20.6%, 12.3% and 15.5%, respectively. Co-occurrences of the disorders were noted. Zuckerman [66] investigated the prevalence of headache and stomachache in a sample of 308 3-year-

old children. An interview with the mother of each child was used to determine if the child experienced either headache or stomachache within the previous 4 weeks. The interview procedure was not documented and no attempt was made to establish the severity of pain experienced by each child. Approximately 3% of the sample suffered from recurrent stomach ache. Twenty-two percent of those suffering from recurrent headache and 9% suffered from recurrent stomachache also suffered from recurrent headache. Aro et al. [3] documented the prevalence of various psychosomatic symptoms in a group of 2246 adolescents (1002 males, 999 females; aged 14 to 15 years). All participating subjects completed a questionnaire about the occurrence of psychosomatic symptoms. The prevalence of recurrent headache pain was 7.2% in males and more than twice as high in females at 14.9%. Of all subjects who completed a questionnaire, approximately 2.5% reported abdominal pain and 1% reported heartburn. Data on the prevalence of co-occurrence of different pains in the same individual are particularly important in understanding if some individuals are pain prone. Unfortunately, these data are extremely limited.

Pain in clinical settings In this section of our review, we will include all studies in which the sample under investigation is drawn from a clinical population. A clinical population can be defined as any group that has sought medical attention for any reason other than a routine or mandatory physical examination. As previously mentioned, pain is a symptom which often prompts many but not all people to seek medical attention [lo]. However, the reason for seeking medical attention or for admission to hospital described in each of the studies reviewed may not have been due to pain. By studying a clinical population, an epidemiological study can accomplish two important things. First, it is possible to determine how well pain is being managed in hospitals and clinics and which pain management techniques are most successful. Secondly, a better understanding of the pain commonly experienced with various clinical conditions or pain associated with various clinical experiences can be acquired, The studies investigating the incidence and prevalence of pain in children and adolescents in a clinical setting are summarized in Table III. Head pain

As headache pain is quite common during childhood [52], Jay and Tomasi 1241evaluated the type of headache severe enough to warrant a referral by a pediatrician to the Neurology Unit at Memorial Children’s Hospital in

25x TABLE

111

SURVEYS

EXAMINING

THE EPIDEMIOLOGY

OF PEDIATRIC

PAIN IN CLINICAL

POPULATIONS

representativeness

Critical

summary

Disorder

Sample

Sample

headache

n = 53Y patients (aged 5-16 years) referred to a pediatric neurology clinic

No sex distribution of sample given; male : female ratio given according to diagnosis

Patient charts were reviewed retrospectively; 22% of those referred to the neurology clinic were referred because of headache; sex prevalence, family history and associated symptoms were also assessed.

of 146 children originally targetted for study, 129 (70 males, 59 females) patients aged 5-14 years, admitted with closed head injury. were used in final sample

Original sample was well described, but failed to document how the two follow-up groups were divided

Headache prior to injury was not assessed: follow-up was not standardized: measurement of headache was not detailed; time frame for recall at follow-up ranged from 6 months to 1 year; reported that 24.6% suffered from headache 6 years following injury.

post-traumatic headache syndrome

n = I38 (86 males, 52 females), M = 9.2 years

Sampling frame was well documented: subjects were grouped according to severity of head trauma

Headache prior to injury was not assessed; used a validated questionnaire to assess headache; after trauma. reported that 29% suffered from headache and 6.5% suffered from migraine.

postoperative

pain

n = 170 (101 males. 69 females), M = 8 years

Sampling detailed

frame

was not well

Assessed postoperative pain in relation to analgesic and narcotic medication prescribed; method of pain measurement was not validated; despite medication ordered or given, 40% of patients reported moderate or severe pain on operative day.

postoperative

pain

n = 20 (17 males, 3 females), aged IX months to 4 years

Sampling tailed

frame

was well de-

Used movement/vocalization expressive of pain as measurement (high interrater reliability was reported for this method); prevalence of pain was reported in terms of no. of movements/ vocalizations and thus difficult to interpret.

pain in hospitalized patients

of I52 children originally targetted for study, 150 patients, aged 4-14 years, were included in the final sample; sex distribution was not reported

Sampling frame was well detailed, but exact age and sex distribution was not reported

Used accepted pain measurement techniques; 57% of hospitalized sample reported moderate to severe pain; pain became resolved in most of the patients within a period of one month.

pain caused by blood sampling

of lY6 children targeted for study (55.1% male, 44.9% female), 171 were included in sample, aged 3-17 years

Sampling tailed

Used accepted and validated pain measurement techniques; depending on the pain measure used, 36-64s of children aged 3-h years experienced moderate to severe distress from blood sampling: 52% of children 7-17 reported pain from blood drawing.

Reference I. Head pain

Jay and Tomasi

Lanser

[24]

post-closed head jury headache

et al. [28]

Lanzi et al. [ZY]

2. Pain WI hospitalised Mather

Taylor

Johnston

Fradet

and Mackie

[56]

et al. [25]

et al. [I81

in-

patients

[3I]

frame

was well de-

259 TABLE

III (continued)

Reference

Disorder

Sample

Sample

representativeness

Critical

summary

3. Cancer pain Miser et al. [39]

cancer

related

pain

n = 139 (60 inpatients, 79 outpatients), M = 16 years

Age range and distribution, and sex distribution of subjects not reported, sampling frame was not well detailed

Used an accepted and validated visual analogue scale to measure pain; pain was present in 54% of inpatients and in 26% of outpatients, when analyzing the 356 ‘pain visits.’

Miser et al. [40]

cancer

related

pain

n = 92 nosed, months M = 16

Age and reported. setting

sex distribution not Not from a typical

Used an accepted and validated visual analogue scale to measure pain; 72 of 92 patients were experiencing pain prior to starting treatment. Pain resolved shortly after treatment began. Many with moderate or severe pain were not given analgesics.

McGrath

cancer

related

pain

n = 77 (45 males, 32 females), aged 2-19 years

Sampling frame was well documented; included all patients in a pediatric cancer clinic

Assessed pain from disease, diagnostic procedures, and treatments; reporting moderate to severe pain was 37% from disease, 41% from chemotherapy, 78% from bone marrow aspirations and 61% from lumbar punctures.

n = 427 (178 males, 249 females), age range lo-70 years

Sampling frame was well documented, included all patients admitted to hospital because of unexplained abdominal pain

Severity of pain is not documented, rate of admission to hospital for abdominal pain was reported at 9% for males and 16% for females, for the age group lo-19 years.

n = 30, aged 15at time of follow-up II. n=30, aged 1528 years at time of follow-up

Sample characteristics were given, only studies of abdominal pain where no organic cause could be isolated

Compared 2 series of RAP patients: (I) treated with reassurance vs. (II) no treatment; compared in each numbers of spontaneous recoveries and frequency of other disorders; pain persisted in 36% of group I and 40% of group II.

all children aged 15 years or less, referred to an orthopedic clinic during a 6-year period

Number in sample was not reported, nor was age or sex distribution, only age distribution of back pain sufferers was reported

In the (i-year period, approx. 2% of referrals were because of back pain, severity of pain is not documented but showed that back pain increases dramatically with age.

back

n = I42 top athletes (117 males, 26 females), age range 14-25 years

Sampling frame was well documented; although not a clinical group by definition, these subjects represent a highly specialized group

Assessed back pain in relation to radiologic changes in thoracolumbar spine of top athletes, prevalence of back pain tmoderate-severe) in this group was 64.8%.

syn-

n = 1249 (562 males, 687 females) students of 7 Australian music schools

Age range of sample was not reported; included all students of 7 of Australia’s I1 performing music schools

Data was collected using an interview between the students and staff of the schools and the author; the syndrome was present in 11.1% (76) females and 7.1% (40) males; severity was assessed using a graded 5level scale; the disorder was most prevalent in woodwind players and least prevalent in brass players.

et al. [34]

newly diagage range 6 to 24 years. years

4. Abdominal pain Rang et al. [47]

‘undiagnosed’ dominal pain

Apley and Hale [I]

recurrent pain

ab-

abdominal

I.

24 years

5. Back pain

Turner

et al. [58]

back pain

Sward et al. (551

sports pain

Pry ]l91

overuse drome

related

injury

260 TABLE

III (continued)

Reference

Disorder

Fry and Rowley [20]

overuse drome

injury

syn-

Sample

Sample representativeness

Critical

n = 169 (62 males, 107 females) music students, age range 7-19 years; n = 348 (167 males, 181 females) regular students, age range lOI9 years

Sampling frame was well detailed; included all students of one regular and one music school in Australia

Music students were asked to report pain solely from music whereas regular students were asked to report pain from all sources; questionnaire was not validated and no attempt was made to assess severity of pain; lifetime prevalence was 71% in music students and 50% in regular students

summary

out of 336 children targeted, used 225 in sample t I59 males, 166 females). age range 2- 18 years

Sampling tailed

Children targeted during l2month period represented 0.6% of all visits to the emergency department; questionnaire was pretested and considered valid; 90% (288 patients) rated pain as either ‘moderate’ or ‘severe.’

7. Chest puin Rowe et al. 1501

chest pain

Chicago. During a period of 1 year, 539 referrals aged 5- 16 years were examined and categorized according to the type of disorder for which they were referred. According to standard diagnostic criteria for migraine and headache [46], 22% of all patients were referred because of headaches. Almost half of these children suffered from migraine. The relationship between headache and closed head injury was investigated in 2 studies [28,29] which followed children after admission to hospital with a closed head injury. Both found that about 25% of children had headache several years after injury. Control groups were not used and these rates are not clearly above those reported in similar age ranges. Consequently, no relationship between closed head injury and long-term headache can be inferred. Pain in hospital patients Mather and Mackie [31] measured the incidence of pain in a group of 170 (101 males, 69 females; M = 8 years) children recovering from surgery. Each patient was interviewed between 2 and 4 times each day for 2-3 days, depending on the postoperative recovery time spent in hospital. The method of pain measurement had been previously used but had not been validated. Although a postoperative narcotic or nonnarcotic analgesic was ordered for 84% of the sample, approximately 40% of patients reported moderate or severe pain on the operative day following surgery. A similar study by Taylor [56] determined the incidence of postoperative pain in a sample of 20 children (17 males, 3 females), the majority of whom were toddlers and preschoolers, aged between 18 months and 4 years. None of the patients in the study received

frame

was well de-

any postoperative medication. Pain was measured by the number of vocalizations and/or movements indicative of pain. Although a high inter-rater reliability was reported for this method, it had not been validated at the time of the study. Not surprisingly, it was shown that every subject exhibited some form of movement or vocalization indicative of pain, but there was no attempt to relate the obtained measure to pain intensity or severity. Furthermore, a baseline measure of similar vocalizations and movements was not taken prior to surgery. Taylor concluded that the majority of movements consisted of expressions of pain, frustration and rage, but this conclusion was largely speculative and empirically unfounded. To assess the pain commonly experienced by hospitalised children, Johnston et al. [25] studied 152 children, aged 4-14 years, hospitalised in the pediatric hospital of McGill University. Of these, 150 were included in the final sample. Using several techniques to measure pain, all of which had been previously validated, approximately 57% of the sample reported severe pain. At a 3-week follow-up, 84% of the original sample was contacted and, of these, pain had completely resolved in 68.3%. Nineteen percent of those contacted were experiencing pain daily and 12.7% were experiencing pain weekly. Hospital procedures can also be a source of pain and distress for children and adolescents. Fradet et al. [18] studied 196 children (55.1% male, 44.9% female) at the Children’s Hospital of Eastern Ontario in a prospective survey of pain experienced during blood sampling. Of these, 171 children and adolescents aged 3-17 years, who underwent venepuncture for blood sampling, were included in the final sample. Using

261

validated measures, subjects were questioned about the pain and anxiety experienced during the blood sampling and were observed immediately before and after the procedure. Depending on the pain measure used, 36-64% of children aged 3-6 years experienced moderate to severe distress. Children in this age group also exhibited significantly more distress behaviours than children over the age of 7 years. Previous experience with either venepuncture or other needle procedures did not significantly affect distress.

was assessed using a thorough physical examination. For a study period of 2 years, 427 patients (178 males, 249 females; aged lo-80 years) were discharged from hospital with this disorder. Of this group, 124 patients were between the ages of 10 and 19 years (47 males, 77 females). In this youngest age group, the incidence of ‘undiagnosed’ abdominal pain discharges for the hospital population included in the Oxford Linkage Study Area (1962-1963) was 9% for males and 16% for females.

Cancer pain

Back pain

Cancer pain including pain directly resulting from the disease itself, treatments or invasive diagnostic procedures has been reported in 3 studies. The first was a survey of the prevalence of pain in 139 pediatric cancer patients (60 inpatients, 79 outpatients) at the American National Cancer Institute [39]. An accepted and validated visual analogue scale was used to measure pain. The source of subjects but not the age range, age distribution or sex distribution was reported. Pain was present in 26% of outpatients and in 54% of inpatients. Although the source of the pain occurring as a result of either disease, treatments or diagnostics was reported, the severity of pain from different sources was not. The second survey [40] reported on pain as a presenting symptom in children and young adults with newly diagnosed malignancy. Seventy-two of the 92 respondents were experiencing pain prior to diagnosis. Almost all had pain remission within 2 weeks of the beginning of cancer treatment. A limitation of both of these studies is that the patients of the National Cancer Institute are not typical of the patients of most pediatric cancer agencies. The final study [34] surveyed 77 outpatients (45 males, 32 females; aged 2-19 years) from a pediatric oncology clinic regarding pain from disease, treatments and diagnostic procedures. Pain ratings were obtained using a validated visual analogue scale. However, all reports were retrospective and no inpatients were included. Thirty-seven percent of the children experienced moderate or severe pain from the disease. Of those receiving chemotherapy treatment, 41% reported moderate to severe pain. Seventy-eight percent of children receiving bone marrow aspirations and 61% of those receiving lumbar punctures reported moderate to severe pain. More studies with a wider range of subjects are needed.

As noted in the review of non-clinical studies, this pain problem has not been widely investigated. Turner et al. [58], reviewed the charts of all patients aged 15 years or less during the period 1978-1984, who were referred to an orthopedic clinic. Little information about the sample was provided, but during the 6-year period of study, 61 children were referred to the clinic. This represented less than 2% of referrals for this age group. In over half of these subjects there was evidence of serious spinal disease. Back pain and its relationship to changes in the thoraco-lumbar spine of athletes was studied by Sward et al. [55]. Although as a group athletes do not correspond to a clinical group by definition, because of the specialized nature of the subject pool and the excessive demands put on the back in this group, this study was included in the clinical section of the review. One hundred and forty-two top athletes (116 males, 26 females; aged 14-25 years) active in their sport since the age of 10 years agreed to participate in the study. All subjects completed a questionnaire about the occurrence of back pain and had lower thoracic and entire lumbar spine X-rays. Moderate to severe back pain was reported by 64.8% of subjects and radiologic abnormalities were found in 36-55% of subjects, depending on their sport. The authors did not note whether back pain persisted when excessive athletic activity was discontinued.

Abdominal pain

Rang et al. [47] recorded the incidence of admission to hospital due to abdominal pain in which no definite organic explanation could be cited as the primary cause. This was defined as ‘undiagnosed’ abdominal pain and

Overuse injury syndrome

Overuse injury syndrome refers to those changes brought about in the muscles and ligaments from excessive use, causing pain, loss of function (disability) and tenderness in the affected structures [19]. Two surveys [19,20] have examined overuse injury syndrome in school children who were musicians. Fry [19] interviewed and examined the students of 7 of Australia’s performing music schools (enrolment 1249) and found that 9.3% (116) of the children were affected. The severity of injury was graded using a 5-level anchored scale. Woodwind players were most severely affected, while brass players were least likely to report a problem. Hands were most affected but neck, spine and facial muscles were also involved in some cases. In all

262

groups, females were more likely than males to be suffering. Onset of symptoms was almost always correlated with increase in length and intensity of practice. Students with this syndrome were said to show psychological distress. Although the population was clearly defined, it is not clear if all children were examined. Fry noted that the estimated prevalence is probably low because of the motivation not to report pain. The psychometric properties of the interview and severity rating were not established. Determination of psychological distress was by clinical impression. In a questionnaire study, Fry and Rowley [20] surveyed all students at a music school and a regular school to calculate lifetime prevalence of pain in the hands or arms apparently due to playing music or other activities such as sporting activities, writing or lifting. Children in the music school were asked to report pain from music whereas the children in the regular school were.asked about pain from all sources. Prevalence was 71% for the music school students. The regular students were most afflicted by writing pain, with females most likely to have pain. Cello students had the highest prevalence of pain and brass players the least. The representativeness of the sample is unknown. The psychometric properties of the questionnaire were not known and severity was not measured. Pain from overuse is a problem for many musicians. Future studies should use better measures and appropriate sampling. Chest pain

Although pediatric chest pain is a relatively common complaint in primary care [50], neither its incidence nor its prevalence has been documented. The characteristics of children presenting to a pediatric emergency department with chest pain was studied by Rowe et al. [501. The researchers initially examined 336 consecutive children presenting with the complaint of chest pain to the emergency department of a children’s hospital. This represented 0.6% of all visits to the department during a 12-month period. Data were collected using a questionnaire designed specifically for this study, which had been pretested using children experiencing a number of different pain problems. Those children 12 years of age and older completed the questionnaire themselves while for those children younger than 12 years, parental reports were used. Of the original 336 children enrolled in the study, 11 left the hospital before being physically examined leaving a sample of 325 children (159 males, 166 females; aged 2-18 years). After physical examination, the patient was classified according to 1 of 6 diagnostic categories defined according to type of presenting symptoms: chest wall (90 patients), miscellaneous (68 patients), pulmonary (62 patients), traumatic (49 patients), idiopathic (39 patients) and psychogenic (17 patients). Only

10% (31 patients) of the final sample rated their chest pain as ‘mild,’ whereas the remaining 90% (288 patients) rated it as either ‘moderate’ or ‘severe,’ defined as either sometimes or frequently interfering with activity, respectively. The researchers concluded that although chest pain is a common complaint in children, it is a relatively infrequent cause for seeking emergency medical care and usually not indicative of any serious medical problem.

Discussion

As noted in the majority of the studies which have come under review, there are common design and measurement problems (see Table I) which hinder their interpretation. Often researchers have the misconception that because a particular questionnaire is frequently used [e.g., 61, it will provide accurate and valid data. Scales have been developed for the clinical measurement of pain in children [see reviews 5,32,35] but no work has been done on survey instruments. As previously mentioned the status of measures of disability is even poorer. Studies are needed to validate survey measures of pain and disability for children. In the absence of sophisticated measurement techniques, many studies have made no attempt to quantify the intensity or even frequency of pain complaints. Although such studies may be informative, they are severely limited as the severity and extent of pain are often as important as the occurrence of the pain. Unfortunately, for many disorders, there are no widely accepted criteria for deciding whether or not a specific amount of pain is ‘normal’ or should be considered a case. Epidemiologic studies can provide the data to help make these determinations. The power of epidemiology to assist in the understanding of the etiology of pain and resulting disability and handicap has not yet been harnessed. For example, families are often thought to play a role in chronic pain and its sequelae. However, aggregation of pain problems among family members has not been appropriately investigated and the mechanisms of transmission of pain problems within families are unknown. Although several studies have questioned respondents about pain experiences in family members [e.g., h-81, no studies have directly questioned different members of the same family about similar pain experiences. Until recently, pain in children and adolescents has been ignored and largely misunderstood [33]. As research in this field continues to expand and become more sophisticated, the work that has been done to date must be critically evaluated. From this review, it is clear that the epidemiology of pain in children and adolescents is still relatively undocumented.

263

References 1 Apley, J. and Hale, B., Children 2

3

4 5 6 7 8 9 10 11

12

13 14

15

16

17

18

19 20 21

22 23

24 25

with abdominal pain: how do they grow up?, Br. Med. J., 3 (1973) 7-9. Apley, J. and Naish, N., Children with recurrent abdominal pains: a field survey of 1000 school children, Arch. Dis. Child., 33 (1978) 165-170. Aro, H., Paronen, 0. and Aro, S., Psychosomatic symptoms among 14-16 year old Finnish adolescents, Sot. Psychiatr., 22 (1987) 171-176. BalaguC, F., Dutoit, G. and Waldburger, M., Low back pain in schoolchildren, Stand. J. Rehab. Med., 20 (1988) 175-179. Beyer, J.E. and Wells, N., The assessment of pain in children, Pediatr. Clin. N. Am., 36 (1989) 837-854. Bille, B., Migraine in schoolchildren, Acta Pediatr. Stand. Suppl., 51 (1962) 1-151. Bille, B., Migraine in childhood and its prognosis, Cephalalgia, 1 (1981) 71-75. Bille, B., Migraine in childhood, Panminerva Med., 24 (1982) 57-62. Collin, C., Hockaday, J.M. and Waters, W.E., Headache and school absence, Arch. Dis. Child., 60 (1985) 245-247. Crook, J., Rideout, E. and Browne, G., The prevalence of pain complaints in a general oopulation, Pain. 18 (1984) 299-314. Dalsgaard-Nielsen, T., Enberg-Pedersen, H. and Holm, H.E., Clinical and statistical investigations of the epidemiology of migraine, Dan. Med. Bull., 17 (1970) 138-148. Davison, I.S., Faull, C. and Nicol, A.R., Research note. Temperament and behaviour in six-year-olds with recurrent abdominal pain: a follow up, J. Child Psychol. Psychiat., 27 (1986) 539-544. Deubner, D.C., An epidemiologic study of migraine and headache in lo-20 year olds, Headache, 17 (1977) 173-180. Egermark-Eriksson, I., Ingervall, B. and Carlsson, G.E., Prevalence of mandibular dysfunction and orofacial parafunction in 7-, ll- and 15-year-old Swedish children, Eur. J. Orthod., 3 (1981) 163-172. Fairbank, J.C.T., Pynsent, P.B., Van Poortvliet, J.A. and Philips, H., Mechanical factors in the incidence of knee pain in adolescents and young adults, J. Bone Joint Surg. (Br.), 66 (1984) 685-693. Fairbank, J.C.T., Pynsent, P.B., Van Poortvliet, J.A. and Philips, H., Influence of anthropometric factors and joint laxity in the incidence of adolescent back pain, Spine, 9 (1984) 461-464. Faull, C. and Nicol, A.R., Abdominal pain in six-year-olds: an epidemiological study in a new town, J. Child Psychol. Psychiatr., 27 (1986) 251-260. Fradet, C., McGrath, P.J., Kay, J., Adams, S. and Luke, B., A prospective survey of reactions to blood tests by children and adolescents, Pain, 40 (1990) 53-60. Fry, H.J.H., Prevalence of overuse (injury) syndrome in Australian music schools, Br. J. Industr. Med., 44 (1987) 35-40. Fry, H.J.H. and Rowley, G.L., Music related upper limb pain in schoolchildren, Ann. Rheum. Dis., 48 (1989) 998-1002. Golub, L.J., Lang, W.R. and Menduke, H., The incidence of dysmenorrhea in high school girls, Postgrad. Med., 23 (1958) 38-40. Harris, L. and Associates, Inc., The Nuprin Pain Report. Study Nr. 851017, New York, 1985. Hughes, E.L. and Cooper, C.E., Some observations on headache and eye pain in a group of schoolchildren, Br. Med. J., 1 (1956) 1138-1141. Jay, G.W. and Tomasi, L.G., Pediatric headaches: a one year retrospective analysis, Headache, 21 (1981) 5-9. Johnston, C.C., Jeans, M.E., Gray-Donald, K. and Abbott, F.V., A survey of pain in hospitalised patients aged 4-14 years, in press.

26 King, H.A., Back pain in children, Pediat. Clin. N. Am., 31 (1984) 1083-1095. 27 Klein, J.R. and Litt, I.F., Epidemiology of adolescent dysmenorrhea, Pediatrics, 68 (1981) 661-664. 28 Lanser, J.B.K., Jennekens-Schinkel, A. and Peters, A.C.B., Headache after closed head injury, Headache, 28 (1988) 176-179. 29 Lanzi, G., Balottin, V., Borgatti, R., DeAgostini, G.B., Pezzotta, S. and Spanu, G., Late post-traumatic headache in pediatric age, Cephalalgia, 5 (1985) 211-215. 30 Linet, MS., Stewart, W.F., Celentano, D.D., Ziegler, D. and Sprecher, M., An epidemiologic study of headache among adolescents and young adults, J. Am. Med. Assoc., 261(1989) 2211-2216. 31 Mather, L. and Mackie, J., The incidence of post-operative pain in children, Pain, 15 (1983) 271-282. and Treat32 McGrath, P.A., Pain in Children. Nature, Assessment ment, Guilford, New York, 1990. 33 McGrath, P.J., Paediatric pain: a good start, Pain, 41 (1990) 253-254. 34 McGrath, P.J., Hsu, E., Cappelli, M., Luke, B., Goodman, J.T. and Dunn-Geier, J., Pain in pediatric oncology: a survey of an outpatient clinic, J. Psychosoc. Oncol., 8 (1990) 109-124. 35 McGrath, P.J. and Unruh, A.M., Pain in Children and Adolescents, Elsevier, Amsterdam, 1987. P.J., Mathews, J. and Pigeon, H., Assessment of pain 36 McGrath, in children. In: M.R. Bond, J.E. Charlton and C.J. Woolf (Eds.), Pain Research and Clinical Management, Vol. 4, Proc. of the VI World Congress on Pain, Elsevier, Amsterdam, 1990, pp. 509-526. 37 Merskey, H., Classification of chronic pain: descriptions of chronic pain syndromes and definitions of pain terms, Pain, Suppl. 3 (1986) Sl-S8. 38 Mikail, S.F. and Von Baeyer, C.L., Pain, somatic focus, and emotional adjustment in children of chronic headache sufferers and controls, Sot. Sci. Med., 31 (1990) 51-59. 39 Miser, A.W., Dothage, R.A., Wesley, R.A. and Miser, J.A., The prevalence of pain in a pediatric and young adult cancer population, Pain, 29 (1987) 73-83. 40 Miser, A.W., McCalla, J., Dothage, J.A., Wesley, M. and Miser, J.S., Pain as a presenting symptom in children and young adults with newly diagnosed malignancy, Pain, 29 (1987) 85-90. 41 Naish, J.M. and Apley, J., ‘Growing Pains’: a clinical study of non-arthritic limb pains in children, Arch. Dis. Child., 26 (1951) 134-140. 42 Nilner, M. and Lassing, S., Prevalence of functional disturbances and diseases of the stomatognathic system in 7-14 years olds, Swed. Dent. J., 5 (1981) 173-187. 43 0ster, J., Recurrent abdominal pain, headache and limb pains in children and adolescents, Pediatrics, 50 (1972) 429-436. 44 0ster, J. and Nielsen, A., Growing pains: a clinical investigation of a school population, Acta Paediatr. Stand., 61 (1972) 329-334. 45 Passchier, J. and Orlebeke, J.F., Headaches and stress in schoolchildren: an epidemiological study, Cephalalgia, 5 (1985) 167-176. 46 Prensky, A.L. and Sommer, D., Diagnosis and treatment of migraine in children, Neurology, 29 (1979) 506-510. 47 Rang, E.H., Fairbairn, A.S. and Acheson, E.D., An enquiry in the incidence and prognosis of undiagnosed abdominal pain treated in hospital, Br. J. Prev. Sot. Med., 24 (1970) 47-51. 48 Richardson, G.M., McGrath, P., Cunningham, S.J. and Humphreys, P., Validity of the headache diary for children, Headache, 23 (1983) 184-187. 49 Riolo, M.L., Brand& D. and TenHave, T.R., Associations between occlusal characteristics and signs and symptoms of TMJ dysfunction in children and young adults, Am. J. Orthod. Dentofacial Orthop., 92 (1987) 467-477. 50 Rowe, B.H., Dulberg, C.S., Peterson, R.G., Vlad, P. and Li, M.M., Characteristics of children presenting with chest pain to a

264

51 52 53

54

55

56 57 58 59

pediatric emergency department, Can. Med. Assoc. J.. 143 (1990) 388-394. Sillanpaa, M., Prevalence of migraine and other headache in Finnish children starting school, Headache, 15 (1976) 288-290. Sillanpaa, M., Prevalence of headache in prepuberty, Headache, 23 (1983) 10-14. Sillanpaa, M., Changes in the prevalence of migraine and other headaches during the first seven school years, Headache, 23 (1983) 15-19. Sparks, J.P., The incidence of migraine in schoolchildren: a survey by the Medical Officers of Schools Association, Practitioner, 221 (1978) 407-411. Sward, L., Hellstrom, M., Jacobsson, B. and Peterson, L., Back pain and radiologic changes in the thoraco-lumbar spine of athletes, Spine, 15 (1990) 124-129. Taylor, P.L., Post-operative pain in toddler and preschool age children, Matern. Child Nurs. J.. 12 (1983) 35-50. Teperi, J. and Rimpela, M., Menstrual pain, health and behavior in girls, Sot. Sci. Med., 29 (1989) 163-169. Turner, P.G., Green, J.H. and Galasko, C.S.B., Back pain in childhood, Spine, 14 (1989) 812-814. Vahlquist, B., Migraine in children, Int. Arch. Allergy, 7 (1955) 348-355.

60 Varni. J.W., Thompson. K.L. and Hanson. V., The Varni/ Thompson pediatric pain questionnaire. 1. Chronic musculoskeletal pain in juvenile rheumatoid arthritis, Pain, 28 (1987) 27-3X. 61 Waters, W.E., Community studies of the prevalence of headache, Headache, 9 (1970) 178-186. 62 Waters, W.E., Migraine and symptoms in childhood: bilious attacks, travel sickness and eczema, Headache, 12 (1972) 55-61. 63 Widholm. 0.. Dysmenorrhea during adolescence, Acta Obstet. Gynecol. Stand., 87 (1979) 61-66. 64 Wigdorowicz-Makowerowa, N.. Grodzki, C., Panek, H., Maslanka, T., Plonka, K. and Palacha, A., Epidemiologic studies on prevalence and etiology of functional disturbances of the masticatory system, J. Prosthet. Dent.. 41 (1979) 76-82. 65 World Health Organization, International Classification of lmpairments, Disabilities and Handicaps. World Health Organization, Geneva, 1980. 66 Zuckerman, B., Stevenson, J. and Bailey, V., Stomachaches and headaches in a community sample of preschool children. Pediatrics, 79 (1987) 677-682.