EVALUATION OF TWO QUESTIONNAIRES TO DETERMINE EXPOSURE TO RISK

Evaluation of two questionnaires to determine exposure to risk factors for non-specific low back pain in Mallorcan schoolchildren and their parents

Background: A pilot study was undertaken to test the methodology as well as the comprehensibility, validity and reliability of two questionnaires to be used in a study to determine prevalence of low back pain In schoolchildren In Mallorca and their parents. Methods: Fifty students from a school in Palma de Mallorca, aged 13-15 years, were surveyed from September to December 1996, as well as their parents. The questionnaires were distributed to the students by the study's school coordinator. Evaluation of the questionnaires was done through the test-retest method, the test through a self-administered, written version and the retest through an interview. Questions were asked on presumed risk factors for low back pain and on topics associated with its characteristics. Results: The system designed for the data collection phase was successful. Difficulties with comprehension centred mainly around two questions: sports and alcohol intake. Validity was only assessed on two student questions (academic problems and ever/never diagnosis of scoliosis) and the validity measures used were concordance of students' and parents' responses and concordance of students' responses with the gold standard (academic and medical records). With respect to reliability there was a good test-retest correlation for each subject, except for students' hours of television watching, associated leg pain and problems with schoolwork (p=0.013, 0.043, and <0.001, respectively); and in parents' problems with schoolwork in their child (p<0.0001). Conclusion: Other than the necessity of making some minor adjustments to the questionnaires, it appears that they are adequate for collecting the information necessary for this study.

Keywords: low back pain, pilot study, schoolchildren, validation

N

on-specific or common low back pain (LBP) is defined as pain in the lumbosacral region, usually accompanied by painful limitation of motion and is influenced by strain and the adoption of certain postures, which may be associated with referred pain. When diagnosed it is assumed that there are no underlying conditions such as fractures, spondylitis, direct trauma or neoplastic, infectious, vascular, metabolic or endocrinological-related processes which might cause pain. LBP is attributed to degenerative processes of discs or facet joints, disc protrusion or herniation, muscular strains and other disorders associated with the position or movement of the spine, such as scoliosis * M.T. Gil del Real1, F.M. Kovaa', M. Gestcso1, N Mufraggl 1 , J.M. Dieguez', Balearic Back Pain Group* 1 Fundadcin Kovaa, Madrid and Palma de Mallorca, Spain a- Other members include the following MDs: Silvia Agullar, Onofre Alba, Antonio Bennasar, Teodonco Cabanes, Jaime Canei. Antonio Godoy, Pfa Klapsing, Jose M. Magrinya, Irene Pascual, Felix Pons, Jose J. Ramos, Bartolome Ribas, Miguel Rubl, Antonio M. Sabater, Encamacion Sanfelidano, Rafael Suau.

Comipondenc*: Maria Teresa Gil del Real, MPH, Coordinator Scientific Department Fundadon Kovaa, Plaza Valparaiso 8, 28036 Madrid, Spain, tel. +34 91 3440244, fax +34 91 3441950, e-mail: mtgildelrealakovacs.org

or spondylolisthesis.1 In most cases, however, it is not possible to establish an organic cause for the symptomatology.1"3 LBP is one of the most frequent pathologies in industrialized countries, ' ^ and is actually among those which generate the greatest expense due to health and labour costs.1'6 The majority of episodes of back pain are limited in time and have a fairly good prognosis.'' Nevertheless, there is a tendency to relapse and to have increasingly more intense episodes. • Although it is believed that LBP is infrequent among children and adolescents, population-based studies have demonstrated that they often complain of LBP.1" According to international data, the prevalence of LBP in children and adolescents varies between 7 and 63% 9>11-21 and tliere are data which indicate that the academic performance of children with LBP might be inferior to that of those who do not have this pathology.11 A Danish study found that X-ray changes in the lower spine in adolescents, including scoliosis, are not risk factors for LBP in adults, but that having LBP as an adolescent is associated with an increased risk of having it as an adult, as well as a salary lower than the average popula-

Downloaded from https://academic.oup.com/eurpub/article-abstract/9/3/194/589797 by guest on 15 September 2018

MARIA TERESA GIL DEL REAL, FRANCISCO M. KOVACS, MARIO GESTOSO, NICOLE MUFRAGGI, JOSE M. DlfcUEZ, BALEARIC BACK PAIN GROUP *

Low back pcdn in schoolchildren

METHODS A pilot study was done in order to test the study's methodology as well as the questionnaires before starting the large study in the Mallorcan student population. The objective of the pilot study was twofold. • To test the data collection methodology: how long do the schoolchildren take to fill out the questionnaire and how long do their parents take in returning the completed questionnaire. • To assess comprehensibility, validity of some of the questions and reliability.

Study Population Students from the Cide School in Palma de Mallorca aged 13-15 years were surveyed from September to December 1996. The total study population for the pilot study was N=150: 50 children, as well as their parents or guardians (100), returned the completed questionnaire. This school was chosen for the pilot study because it was the one most generally representative of all Mallorcan schools: its students come from both rural and urban backgrounds, and are of varied socioeconomic status - they come from both rich and poor families. It also offered the right setting for testing the comprehensibility of the questionnaire, since the parents of the students have a mixed educational level. Questionnaire The student questionnaire, referenced on the one used in the Fribourg study,11 collects information on items which, according to the scientific literature, appear to be the most important risk factors for LBP and items associated with its characteristics and/or consequences.

Topics associated with risk factors: - weight and height. —sedentariness: watching television (hours per day, added to hours sitting at school).11 - physical activity involving lower spine: practice of sports (which sports, with what frequency and at what level),11 and mode used for transporting books and materials.5'11 - vibration: use of motor bikes (how much time daily).5 - alcohol intake (kind and amount per week). - smoking (amount per day).26 Topics associated with characteristics and/or consequences of

LBP: - history of back pain (frequency). - history of associated leg pain (Y/N). - visits to physician for back pain (Y/N). - treatment received (kind of treatment). - impediments due to back pain (always/sometimes/ never). - existence of back pain in previous week (Y/N). - diagnosis of scoliosis (ever/never). - number of 'failed' or 'troublesome' courses in last 12 months. It appears that LBP is more frequent in children whose parents have been treated for LBP.11 Therefore, for the study in Mallorca a questionnaire was also designed for the parents (or guardians) of the schoolchildren, which included a question on whether or not they were the biological parents. The questionnaire contained 17 questions on their (parents' or guardians') educational level, motor bike use, practice of exercise or sports, lifetime history of LBP, treatment received, disability due to LBP, ever/never diagnosis of scoliosis, recent history of LBP (in previous week), alcohol intake and smoking and questions relating to the child's academic performance, complaining of back pain or being diagnosed as having scoliosis. Data collection The students in the pilot study were randomly selected and stratified by sex. The study coordinator (a schoolteacher) first explained the objectives of the survey to the students. After the explanation the students were given two sets of questionnaires: one for themselves, to be filled out then and there and two questionnaires for each of their parents or guardians to be filled out at home and returned to the coordinator by the student When the test was administered, neither the students nor their parents knew that there would be a retest. Two weeks after the self-administration of the questionnaires, the coordinator again administered the questionnaire to the students at school - this time verbally - and recorded their responses. The-coordinator then administered the questionnaire to the parents or guardians by means of a telephone interview. Thus, each of the subjects (students and their parents or guardians) completed the questionnaire twice within a two week period: once by selfadministration and once by interview. It was felt that a two week interval between test and retest was necessary


tion.22 A study of schoolchildren in Fribourg, Switzerland found that parental history of LBP, competitive activity in some sports and a great amount of time spent watching television increased the risk of LBP among children.11 Additionally, other studies have found that its existence is associated with certain lifestyle habits and lack of fitness.11'12'23~25In light of these findings, a questionnaire is more reliable than X-rays for collecting information on the principal risk factors for pathology of the back. Thus, the questionnaire that was tested in the present pilot study was mainly concerned with questions regarding both known and presumed risk factors for LBP (see Methods section below). In addition, there were questions on topics associated with the characteristics of LBP. In order to determine the prevalence of LBP in schoolchildren in Mallorca and their parents, as well as exposure to potential risk factors, a study is going to be carried out on all schoolchildren who are residents on the island and aged 13-15 years. The expected study population is 14,400 subjects. The instruments of measure will be three self-administered questionnaires: one to be filled out by the student in the classroom and the other two to be filled out by each of the parents (or guardians) at home, so that there will be a total of 43,200 questionnaires.

EUROPEAN JOURNAL OF PUBLIC HEALTH VOL. 9 1999 NO. 3

Statistical analysis Frequency and variability of responses in the test and retest versions of the questionnaire were analysed by means of the Wilcoxon test. RESULTS Assessment of study methodology The system designed for the data collection phase of the study appears to be successful. Coordination between die investigators, school, school coordinator, students and parents worked well. The majority of students returned the questionnaire after 10 minutes, and none took longer than 15 minutes to complete it. The majority of the parents returned die completed questionnaires to the school after two days and only two of diem took longer than four days. Assessment of comprehension In die oral version of the questionnaires, none of the students asked for clarification of any question and 12% of the parents did so. However, in die written version there was a very high percentage of missing values in questions related to sports: 68-96% in students and 3 78-100% in parents. Missing values for sports were

dramatically reduced in die oral retest version (0% for students and 4-12% for parents). In addition to sports, die great majority of questions that were not well understood referred to alcohol intake and die majority of unanswered ones (mainly by die fadiers) had to do with time spent daily on a motorcycle (table I). Assessment of validity Regarding academic performance, diere was a low correlation in die test in responses to 'failed' and 'troublesome' courses between students and dieir parents (p=0.092/0.094). In the retest there was an acceptable one between students and their modiers (p=0.213), and a good one between students and dieir fadiers (p=0.602). Regarding die question on whetiier die student had ever been diagnosed as having scoliosis, diere was high concordance on the retest (modiers p=O.7O5 and fadiers P =O.479)(wWe2). Assessment of reliability In die children diere was a good test-retest correlation, except in die following items in which die retest scores were lower hours of television watching, LBP with associated leg pain and problems widi schoolwork (p=0.013, 0.043 and <0.001 respectively). In die parents, bodi modier and fatiier also had a good test-retest correlation (p>0.0833), except in relation to problems with schoolwork in their child (p<0.0001) (table 3).

DISCUSSION This pilot study shows diat die methodological design of die study for determining exposure to risk factors for LBP in Mallorcan schoolchildren is adequate and feasible. Additionally, die questionnaire has adequate comprehensibility (after adjustments to specific questions), validity and test-retest reliability. The comprehensibility of die questionnaire proved to be acceptable, widi die exception of a few questions. The question regarding sports was formulated in die form of a diagram, widi blanks to be filled in indicating type of sport, level of competitiveness and frequency of practice. There was confusion regarding die proper way to fill in die diagram and a great number of students did not do it correctly, which led to a large percentage of 'missing' values for diis question (table 1). The misunderstanding was corrected in die oral retest version (where 'missing' values for diis question were 0%) and diat portion of the questionnaire has now been rewritten, so diat it is now clear. The students did not seem to relate LBP to associated leg pain, since diey asked questions about it in die oral version. However, diey did not leave the question unanswered in die written version. There was a pattern observed in die questions left unanswered in the parents' questionnaires. The parents did not understand some of die categories of alcoholic mixed drinks which seem to be popular with die young generation. In addition, in die question regarding sports, it did not seem clear to the parents whether they were being


for two reasons: i) after two weeks the subjects would have forgotten how they answered in the first version and ii) since the retest was oral it was necessary to separate it in time from the written version. The students were given 20 minutes to fill out the questionnaire and the time that they took doing so was recorded. The parents were not given any time period for returning the completed questionnaire, but the date of its return was recorded. The name of each student was substituted with a code in the database to protect privacy, but the pilot study was not done anonymously, since (only) the coordinator knew the students that participated. The reason for this was that the student needed to be identified in order to do the retest and it was also necessary to obtain the parents' telephone number for the same reason. The large study, however, will guarantee anonymity. When the test-retest method was decided upon for assessing reliability, the written/oral method was chosen from several options because it was the most ideal method for testing comprehension of the questionnaire in this case. 27 ' 28 The level of comprehension of the questionnaires was measured by the number of questions left unanswered in the written version (test) and by the number of questions for which clarification was asked in the oral version (retest). Validity was only assessed on the two questions in the students' questionnaire which referred to academic performance and an ever/never diagnosis of scoliosis in the child. The type of validity measures used were concordance of students' and parents' responses in both the test and retest versions and comparison of the student's responses with the gold standard: academic and medical records (all schoolchildren have a medical examination for scoliosis). Reliability of the questionnaire was assessed through the test-retest method previously described.

Low back pain tn schoolchildren

have been omitted and the sentence construction has been changed in the question regarding sports so that it is now clear that they are being asked about themselves. Validity was only assessed for two questions in the student questionnaire (asked also of the parents regarding their child): number of 'failed' or 'troublesome' courses in the previous 12 months and ever/never diagnosis of scoliosis by a physician. There were two validity measures used: concordance between the student's and parent's response and concordance of student's response with the gold standard. There was low correlation between the child and both parents in the test version of the academic question, although in the retest it was adequate. This is probably due to the fact that die students (the boys in

Table 1 Ranges of % missing values per interview (N=150; 50% males and 50% females) Variables

Students' mother Test Retest

Students' father Test Retest

0 0 NA

NA NA

NA NA

NA NA

4

0 NA 0 0 0 0

78-100

4 4

0 NA NA 2

14

0 0 0 0 0

24 8 18 26

0 0 0

0 0 0

0 0 0 0

0 0 0 0 0

Students

Test

Retest

0 0 NA 68-96 NA 2 2 2

Questions related to risk factors Weight Height Age Sports* Practising physical exercise Riding motorcycles Carrying schoolbooks Watching television Smoking

4

Alcohol intake Beer

12

Shot Cocktail Wine Others

10 8 18

Questions related to LBP History of LBP Associated leg pain Visits to physician Treatment received Impediments LBP in previous week Scoliosis diagnosed Failing grades

4

4

14

4 4 NA NA

4 4 4 4 4 6

NA 8

NA NA 12 12 12 12 NA NA 12

16

12

42

12 12 12 12

14 86-98 8 8 NA

42 28 46

0

4

8

4 4 4

4

18

6

14 22 12 10 10 12

0 2 2

4

4 4 4 4 4

12 12 12 12 12 12 12 12

a: See explanation in the Discussion section NA: not applicable; LBP- low back pain

Table 2 Concordance between students and their parents (N-150; 50% males and 50% females) Student/mother Retest Test Schoolwotk problems Scoliosis

Table 3 Intra-subject discordance between test and retest" (N=150; 50% males and 50% females)

Student/father Test Retest

Subject Student

Associated leg pain

P

P

P

P

0.092

0.213 0.705

0.094

0.602

Mother

0.564

0.479

Father

0.034

Variable Watching television Failing grades Failing grades Failing grades

Wilcoxon's p 0.013 0.043 0.0001 0.0001 0.0001

a: Only variables in which there waj discordance. In all othen p>0.0833


asked about their children or themselves and so they left most questions in this category unanswered. For this reason the parents also have a high percentage of'missing' values for the sports question. Several fathers left die question on motorcycle riding unanswered. It appears that this was left blank because they do not ride a motorcycle and so did not bother to fill it out. In the student questionnaire, the children answered this question adequately. Both students' and parents' questionnaires have now been modified to clarify questions which were not clearly formulated: the question in the student questionnaire referring to associated leg pain has been deleted and in that of the parents the 'newer' mixed drinks questions

EUROPEAN JOURNAL OF PUBLIC HEALTH VOL 9 1999 NO. 3

The gold standard used for an ever/never diagnosis of scoliosis was the child's medical record at the school (the result of an examination by the school physician on all schoolchildren). The students' answers on the retest were very accurate and correlated with what the physician found on examination of the child. The test-retest method used to assess reliability proved efficacious, except for one detail: the coordinator who administered the oral version of the questionnaire was a schoolteacher in that school, which probably biased some of the responses by both students and parents. The students had a high correlation on their test-retest scores, except for the three items that had to do with television, schoolwork problems and associated leg pain. In the oral (retest) version it was originally planned to leave out the question regarding back pain in the previous 7 days because it was assumed that it would not be relevant or accurate, but in the end it was asked anyway in order to maintain the consistency of the questionnaire. Surprisingly, the results on this question in the retest were consistent with the test results for both children and parents. This can possibly be interpreted as being a chronic condition in those children that reported having back pain (and in those parents who did so as well). From the results of this pilot study it is apparent that greater emphasis should be placed on the need to answer all of the questions (particularly by the parents), in order

to reduce to a minimum the number of missing values. A system to ensure this in the large study has been devised. Since the oral retest answers by both parents and students proved, in the majority of questions, to be more reliable than the written test ones, it might be suggested that the information for the actual study be collected by means of a personal interview. This option was considered, but was believed to be impractical and difficult since the expected study population is 14,400 subjects. In conclusion, these questionnaires appear to be adequate for answering the formulated questions in a reliable and easily understandable manner. The study was supported by the Kovacs Foundation, the Conselleria de Educaci6, Cultura i Sports del Govern Balear and GESA (Gas y Electricidad, S. A.). The authors thank Francisco Mendoza for construction of the database, Jos£ Ignacio Mfndez for statistical analysis of the data and table design and Dr Federico Balagu£ for his consultation services.

1 Deyo RA, Cherkin D, Conrad D, Volinn E. Cost controversy, crisis: low back pain and the hearth of the public Ann Rev Public Hlth 1991,-12:141-56. 2 Vanharanta H, Sachs BL, Spivey M, et al. A comparison of CT/discography, pain response and radiographic disc height Spine 1988;13(3):321-4. 3 Scientific approach to the assessment and management of activity-related spinal disorders. A monograph for clinicians. Report of the Quebec Task Force on Spinal Disorders. Spine 1987;12:S1-59. 4 Coste J, Paolaggi JB. Critical review of the epidemiology of back pain. Rev Epidemiol Sante Publique 1989;37(4):371-83. 5 Kelsey JL, White AA. Epidemiology and impact of low back pain. Spine 1980;5(2):133-42. 6 Ferrer JL Aspectos socioecon6mlcos de las lumbalgias (Socioeconomic aspects of back pain). Tribuna Medica 1980;871:19. 7 Deyo RA. Conservative therapy for low back pain. JAMA 1983;250(8):1057-62. 8 Horal J. The clinical appearance of low back disorder in the city of Gothenberg, Sweden: comparisons of incapacitated probands with matched controls. Acta Orthop Scand 1969;178(Suppl):1-609. 9 Hirsch C, Jonsson B, Lewin T. Low back symptoms in a Swedish female population, d i n Orthop 1969:63:171-6. 10 Turner PG, Green JH, Galasko CSB. Back pain in childhood. Spine 1989;14:812-4. 11 Balague F, Nordin M, Skovron ML et al. Non specific low back pain among schoolchildren: a field survey with analysis of some associated factors. J Spinal Dlsord 1994;7:374-9. 12 Balague F, Distort G, Waldburger M. Low back pain in schoolchildren. Scand J Rehabil Med 1988;20:175-9. 13 Burton AK, Tillotson KM, Troup JDG. Variation in lumbar sagittal mobility with low back trouble. Spine June 1989;14(6):584-90. 14 Fairbank JCT, Pynsent PB, Poortvliet JA, et al. Influence of anthropometric factors and joint laxity in the incidence of adolescent back pain. Spine 1984;9:461-4. 15 Karvonen MJ, Vrtasalo JT, Komi PV, et al. Back and leg complaints in relation to muscle strength in young men. Scand J Rehabil Med 198O;12:53-9. 16 Mierau D, Cassidy JD, Yong-Hing K. Low back pain and straight leg raising in children and adolescents. Spine 1989;14:526-8. 17 Salminen J. The adolescent back: a field survey of 370 Finnish schoolchildren. Acta Paediatr Scand 1984;315(Suppl):8-122. 18 Salminen J. Low back pain and disability in 14 year old schoolchildren. Acta Paediatr 1992;81:1035-9. 19 Tertti M, Salminen J, Paajanen HEK, et al. Low back pain and disk degeneration in children: a case control MR imaging study. Radiology 1991,-180:503-7.


particular) under-reported their academic problems in the test version. However, in the retest version there were adequate correlations between students and their parents. This was obviously because the questions were being asked by a schoolteacher. The question on scoliosis gave a surprisingly low correlation between students and their mothers (p=0.034) in the test version. This was because some students did not understand the term, since in the retest version there was a much higher correlation. The higher correlation of students with fathers on this question in the test version probably occurred because the fathers did not understand the term either or were not adequately informed about their child's health status. For academic performance the gold standard used was school records: these were utilized to assess accuracy of response. The school records showed that girls were sincere regarding their schoolwork in both the written and oral versions of the questionnaire, whereas boys were only sincere in the oral version. It is assumed that, since the retest was done through a face to face interview and not in writing like the test version, the students reported more truthfully regarding their grades in the oral version. The same pattern was observed for the parents, who also had a high test-retest correlation, except for those items related to schoolwork problems in their child (the retest for the parents was done through a telephone interview with the coordinator). This, again, was due to the fact that the students (only the boys) in the test version reported their grades as being better than they actually were. This bias will be corrected in the large study, since it will be anonymous and so the boys will not feel pressured to misreport their grades.

Low back pain m schoolchildren 25 Bergenudd H, Nilsson B, Uden A, Willner S. Bone mineral content gender, body posture and build in relation to back pain in the middle age. Spine 1989;14<6):577-9. 26 Biering Sorensen F, Thomsen C Medical, social and occupational history as risk indicators for low back trouble in general population. Spine 1986; 11:720-5. 27 Wilson RW, McNeil JM. Preliminary analysis of OECD disability on the pretest of the post census disability survey. Rev Epidemiol Sante Publique 1981;29:469-75. 28 Fries JF, Spitz P, Kraines RG, et al. Measurement of patient outcome in arthritis. Arthritis Rheumat 1980,23:137-45. Received J6 May 1998, accepted 18 February 1999


20 van der Linden SM, Fahrer H. Occurrence of spinal pain syndrome in a group of apparently healthy and physically fit sportsmen. Scand J Rheumatol 1988;17:475-81. 21 Wespl H. Haltungsstorungen, Scheuermansche Krankhert und Schularzt (Postural problems, Scheuermann's disease and school physicians). Soz Praeventlvmed 1969; 14:137-45. 22 Harreby M, Neergard K, Hesselsoe G, Kjer J. Are radiologic changes in the thoracic and lumbar spine of adolescents risk factors for low back pain in adults? Spine 1995;20<21):2298-302. 23 Oliver! M. Correct sitting posture: an important element in the therapy and prevention of chronic back complaints. Ergonomic requirements for correct furniture for sitting. Schwelz Reundsch Med Prax 1988;77(25):706-11. 24 Rothbart BA, Estabrook L Excessive pronation: a major blomechanical determinant in the development of chondromalacia and pelvic list. J Manipulat Physiol Ther 1988;11(5):373-9.

EVALUATION OF TWO QUESTIONNAIRES TO DETERMINE EXPOSURE TO RISK

Recommend Documents