The American Academy of Orthopaedic Surgeons (AAOS) Outcomes Studies Committee, collaborating with the Council of Musculoskeletal Specialty Societies and the Pediatric Orthopaedic Society of North America, developed a pediatric functional outcomes assessment instrument aimed at assessing individuals with a musculoskeletal disorder.1 The Pediatric Outcomes Data Collection Instrument (PODCI) was designed as a questionnaire for children aged 2 to 18 years to collect patient-based data for both clinical assessment of treatment effectiveness and musculoskeletal research by evaluating function and quality of life (QOL). In 2002, Hunsaker et al,2 described the normative values from the general population for the PODCI subscales, Upper Extremity Function (UE), Transfers and Basic Mobility (TBM), Sports and Physical Function (SPF), Comfort, Happiness, and Global. Hunsaker reported that standardized scaled scores for specified age, sex, ethnicity, and comorbidity were available through AAOS. Unfortunately, these scores are not available (personal written communication, K. Hitchcock, 2009). There are few published values for PODCI scoring for assessing outcomes in pathologic processes for age or consideration of health/comorbidity.3–6 With restricted ability to match PODCI scores by PODCI norms for age and sex,7–10 comparative values for comorbidities, age, and sex are needed to evaluate treatment outcomes. Further evaluation is needed to determine whether sex, age, and comorbid conditions influence PODCI outcome scores.
Comparison of pretreatment and posttreatment clinical data, with normative data from the general population, is important to assess whether findings in individuals before treatment are different from population norms or come closer to normative ranges of functioning after treatment. In considering the clinical importance of treatment-outcome studies, Kendall et al11 raised 2 basic questions. First, is the pretreatment to posttreatment change large enough to be considered relevant? Second, can treated individuals be distinguished from “normal individuals” serving as a reference group? Clinical measures address the first question, whereas normative comparisons address issues of severity compared with data from a normal, age-adjusted, and sex-adjusted population. Hunsaker’s group noted a need for scores in different ages and sex to assess for normal values.2
Comorbid or health conditions may also affect changes in PODCI scores or be associated with lower outcomes scores.1,12 The PODCI has been reported in the literature assessing children and adolescents with a wide variety of disorders from scoliosis7,10 to cerebral palsy (CP).13–18 When assessing functional outcomes in children, functional levels may not only be affected by growth and maturity factors, but also by other illnesses or conditions that may impact daily life. The increased focus of the International Classification of Function, Disability, and Health to consider a more holistic picture of the impact of medical conditions on an affected individual has made clear the need to consider other factors. Assessing functional outcomes of participation in the community, as opposed to assessing body structure and function measures in the health care setting, necessitates including other factors influencing function. Without a database of PODCI norms to compare with the PODCI outcomes for children with comorbidities, reported PODCI outcomes are incomplete.
Assessments of QOL using various outcomes instruments have demonstrated patterns of responses that vary with age,1,3,19 sex,3,20 type of scales (physical observation, functional scales, or psychosocial scales),21 a particular condition being assessed,1,22 health of respondent,23 and severity of involvement with the particular condition or conditions.23,24 The purpose of this study was to establish the mean and SD for PODCI subscales on the basis of sex, age, and scores reflecting health conditions in a generalized population. To further explore the PODCI, we hypothesized that sex, age, and health/comorbid conditions would affect PODCI outcome scores.
The PODCI was designed to assess the degree to which a patient’s condition or conditions affect his or her physical and emotional functioning, self-image, and symptom status. Three versions of the PODCI include: a questionnaire for children (completed by parents of children under the age of 11) and 2 surveys for adolescents (1 completed by adolescents aged 11 and older and 1 by a parent of that adolescent). Adolescent responses can be matched and compared with parent proxy surveys. The PODCI subscales comprise the following fields: UE, TBM, SPF, Comfort, Happiness, and Global. The PODCI also includes a comorbidity scale with the following health conditions: arthritis, anorexia or bulimia, asthma, attention or behavioral problems, chronic allergies or sinus trouble, developmental delay, mental retardation, diabetes, epilepsy, hearing impairment, learning problems, vision problems, speech problems, sleep problems, and/or heart problems. The respondent answers 3 health condition questions: (1) “Have you (or your child) ever had the problem?” (CM1); (2) “Do you (or your child) receive treatment for it now?” (CM2); and (3) “Does it limit your (your child’s) activity now?” (CM3). Comorbidity subscales and a comorbidity index (CMIndex) that computes an average of the responses are calculated.1
PODCI Database From AAOS
The PODCI AAOS database reported by Hunsaker2 was obtained (with assistance of James Dewey, Mark Kosinski, John Ware, and Michael Goldberg) to evaluate PODCI outcomes in a generalized population sample, which includes subgroups with and without reported health conditions. The PODCI instrument was 1 of 11 musculoskeletal outcome measures evaluated for reliability and reporting normative data. A panel methodology matched US respondents recruited by the National Family Opinion to the 1998 US Census data on geographical region, age, income, household size, and reported comorbidities.2 The Internal Review Board for this institution reported that this secondary data analysis did not meet the criteria for human subjects’ research.
The AAOS database with complete data on the PODCI parent version was used in this study (n=5300). Children identified as having no current or past history of health conditions were compared with children with a health condition in order to verify that the groups have the same stratification as the 1998 census. The data were taken from the CM1 question asking whether the child had ever had a health condition. A χ2 analysis showed no difference in the stratification of the children with or without health conditions with respect to race, sex, and parent version of the PODCI. The sample does not appear to be further stratified by the subgroup of current limiting comorbidities (CM3).
PODCI subscale scores reflected a typical healthy population with higher scores characteristic of better health. Subscale scores demonstrated significant skewness and kurtosis and were transformed to meet the assumptions for analyses. A 1-way multivariate analysis of variance (MANOVA) was conducted to determine whether there was a significant group difference in the prior health condition versus no prior health condition with respect to the dependent variables of the following parent-reported PODCI subscales: UE, TBM, SPF, Comfort, and Happiness scales. Using the Bonferroni method (with a test at P<0.01) to control for Type I error, an analysis of variance (ANOVA) was conducted on each dependent variable as a follow-up to the MANOVA. The eta squared, η2, in the ANOVA reported the proportion of the variance in the PODCI subscale score explained by the comorbid group. Further analysis of health conditions used an independent samples t test to evaluate whether the means of the PODCI subscales for the 2 independent groups, children with a currently limiting health condition (CM3) versus those children who had an identified health condition in the past (CM1), are significantly different from each other. We also used independent samples t tests to investigate whether PODCI scores differed between the sex of children and whether the sex of children within the subgroups of children with and without a health condition had mean group differences. A multiple line graph of PODCI subscale means by age for children with and without health conditions was constructed to further review the effects of age on the PODCI. As a retrospective database analysis, power was not calculated a priori, and a postevaluation power analysis is not recommended.25 The sample size is considered large compared with other studies of the PODCI. Statistical analysis was performed using SPSS (version 21.0; IBM Corp., Armonk, NY). Significance was set at P<0.05.
The most common heath conditions reported by parents were allergy and sinus, asthma, attention/behavioral problems, sleep disturbances, and learning problems (Table 1). With respect to the evaluation of group differences in health conditions, the MANOVA revealed that the combined PODCI subscales were significantly related to whether children ever had a health condition (F5,5043=59.89, P<0.001, multivariate η2=0.056) (descriptive statistics of the PODCI by age, sex, and comorbidities are reported in Appendices 1–3, Supplemental Digital Contents 1–3, http://links.lww.com/BPO/A18, http://links.lww.com/BPO/A19, http://links.lww.com/BPO/A20). For the MANOVA, the η2 suggests that 5.6% of the variance in the PODCI subscale scores was explained by the independent variable, prior health condition or no prior health condition. Subsequent analyses revealed that children with a current or prior health condition experienced lower scores than children without a health condition on the following subscales: TBM (F1,5047=1.82, P<0.001, partial η2=0.005), SPF (F1,5047=16.69, P<0.001, partial η2=0.018), Comfort (F1,5047=29.749, P<0.001, partial η2=0.027), and Happiness (F1,5047=33.85, P<0.001, partial η2=0.035). Significant differences did not emerge for the UE subscale (F1,5047=0.003, P<0.896, partial η2=0.000). Further examination of the group that ever had a health condition (CM1) compared with the group with a currently limiting health condition (CM3) found that children with a current health condition had significantly lower scores on all PODCI subscales (Table 2). With the evaluation of sex and child health conditions, both boys and girls with health conditions had significantly lower subscale scores except for the PODCI UE subscale (Table 3). In terms of child sex differences, boys had significantly higher scores on the PODCI subscale scores of Comfort and Happiness, and no sex differences in scores were found for the other subscales (Table 4). Through the evaluation of age and health conditions, age effects showed an initial increase in the subscales for groups both with and without health conditions and a varied plateau of scores, with decreases in the pain and happiness scales as age increased (Figs. 1, 2).
Although the PODCI has been reported to be a valid and reliable instrument,1,2 PODCI scores should still be considered within the context of sex, age, and reported comorbid conditions. Further evaluation is needed to determine whether the increase in PODCI scores with age could be based on a natural history of improvement in function over time and how children with comorbidities fit into this growth pattern. The typical rate of increase with age in function in certain diagnoses might be shown to be different from the rate in the general population, implying delays in development.26,27 Gait analysis of children with CP has shown improvement up to a point with a plateau effect and even decline in adolescence.28,29 The QOL indices, comfort and happiness, were significantly affected by the sex of the child and showed a decrease in scores in adolescence. Knowing progression of rates in normative group functions allows comparison of functional progression in groups with musculoskeletal disorders.
Incomplete PODCI responses have been reported in the youngest children (ages 2 to 5), which may be due to the response structure. Daltroy et al1 reported that responses were scored as incomplete, especially on the UE or SPF, because the child had not met a milestone (even if “normal” for that age). When adjusting for other health factors, only the UE scale showed age bias. Controlling for age and comorbid conditions in comparing PODCI scores was strongly recommended.
The level of function on PODCI subscales changes with age, as well as with health conditions versus without health conditions. The finding that the parent’s report of the child ever having had a comorbid condition may affect function suggests that the assessment method is valid. When a parent reports that a child is currently limited by a health/comorbid condition, the child demonstrates more limitation than those ever having had a health condition. This finding indicates that there is discriminative ability inherent in the scales. Although the health conditions may vary in severity and are not necessarily medically diagnosed, there is clearly statistical evidence that the ratings are meaningful.
Consideration of comorbidity scores has been suggested when assessing a patient group using the PODCI.1 Nevertheless, only a few articles have reported on comorbidities.12,14,30 A comparison group will be helpful in assessing children with musculoskeletal conditions, if the comorbidities of those groups are measured and compared with the data reported here. As early as 1970, Feinstein31 noted the importance of considering comorbid conditions when comparing groups of patients or interventions. Further evaluation of the limitations of function in patient groups should include the relationship of the limitation with their musculoskeletal condition or associated medical factors.
The child’s sex may also influence the PODCI score as seen in Sheffler et al’s9 study of prosthetic use and parent-child–reported PODCI scores.9 Sex of the child had an impact on parent and self-scores, with the parents of sons scoring greater on comfort than did the boys themselves. Parents rated their daughters lower on UE function than did the girls themselves. Differences in sex and parent responses based on the PODCI and patient Gross Motor Function Classification System levels have also been shown.20 However, Barnes et al3 reported that sex of a patient was not a significant predictor of the parent PODCI for children with CP.
If PODCI scores are collected on all children with a selected condition having musculoskeletal impact, regardless of treatment, meaningful comparisons between treatment and no-treatment groups can be made. A comparison group helps sort out the impact of coexisting factors, enabling a clear focus on the target condition. For example, in an article on children with CP,32 preoperative and postoperative PODCI scale scores were compared, showing increases in some scale scores. Without knowing what would be expected for changes with age, sex, or comorbidity, it is not possible to report that the multilevel surgery made the difference or whether the changes were due to maturation or changes in the comorbidity level. Using not only a population-based comparative database but also collecting data on age, sex, and comorbidity factors on children without surgical treatment as a comparative group would then allow more appropriate assessment of the impact of treatment.
AAOS and Pediatric Orthopaedic Society of North America recognized the value in developing an instrument that can be used to compare effects across different health conditions or diseases.33 It is beneficial to have not only disease-specific outcomes instruments, but also more generic instruments to allow comparisons across disorders and interventions.34 Either functional decline in certain diagnostic groups or a lack of typical development can be assessed by having population-based data available on a large group of children. Having specific details of age, sex, and comorbidities to compare across conditions is a critical piece of outcomes assessment in the era of the International Classification of Function, Disability, and Health and the holistic evaluation of the child.
The authors thank Erin Kelly, PhD, for her editorial support.
1. Daltroy LH, Liang MH, Fossel AH, et al.. The POSNA pediatric musculoskeletal functional health questionnaire: report on reliability, validity, and sensitivity to change. J Pediatr Orthop. 1998;18:561–571.
2. Hunsaker FG, Cioffi DA, Amadio PC, et al.. The American Academy Of Orthopaedic Surgeons Outcomes Instruments. J Bone Joint Surg. 2002;84-A:208–215.
3. Barnes D, Linton JL, Sullivan E, et al.. Pediatric outcomes data collection instrument
scores in ambulatory children with cerebral palsy: an analysis by age groups and severity level. J Pediatr Orthop. 2008;28:97–102.
4. Damiano D, Abel M, Romness M, et al.. Comparing functional profiles of children with hemiplegic and diplegic cerebral palsy in GMFCS Levels I and II: are separate classifications needed? Dev Med Child Neurol. 2006;48:797–803.
5. Lee KM, Chung CY, Park MS, et al.. Level of improvement determined by PODCI
is related to parental satisfaction after single-event multilevel surgery in children with cerebral palsy. J Pediatr Orthop. 2010;30:396–402.
6. McLeod TCV, Bay RC, Parsons JT, et al.. Recent injury and health-related quality of life in adolescent athletes. J Athl Train. 2009;44:603–610.
7. Ugwonali OF, Lomas G, Choe JC, et al.. Effect of bracing on the quality of life of adolescents with idiopathic scoliosis. Spine J. 2004;4:254–260.
8. Bae DS, Waters PM, Zurakowski D. Correlation of pediatric outcomes data collection instrument
with measures of active movement in children with brachial plexus birth palsy. J Pediatr Orthop. 2008;28:584–592.
9. Sheffler LC, Hanley C, Bagley A, et al.. Comparison of self-reports and parent proxy-reports of function and quality of life of children with below-the-elbow deficiency. J Bone Joint Surg Am. 2009;91:2852–2859.
10. Lerman JA, Sullivan E, Haynes RJ. The Pediatric Outcomes Data Collection Instrument
) and functional assessment in patients with adolescent or juvenile idiopathic scoliosis and congenital scoliosis or kyphosis. Spine (Phila Pa 1976). 2002;27:2052–2057, discussion 2057-2058.
11. Kendall PC, Marrs-Garcia A, Nath SR, et al.. Normative comparisons for the evaluation of clinical significance. J Consult Clin Psychol. 1999;67:285–299.
12. Vitale MG, Levy DE, Moskowitz AJ, et al.. Capturing quality of life in pediatric orthopaedics: two recent measures compared. J Pediatr Orthop. 2001;21:629–635.
13. Damiano DL, Gilgannon MD, Abel MF. Responsiveness and uniqueness of the pediatric outcomes data collection instrument
compared to the gross motor function measure for measuring orthopaedic and neurosurgical outcomes in cerebral palsy. J Pediatr Orthop. 2005;25:641–645.
14. Tervo RC, Symons F, Stout J, et al.. Parental report of pain and associated limitations in ambulatory children with cerebral palsy. Arch Phys Med Rehabil. 2006;87:928–934.
15. McCarthy ML, Silberstein CE, Atkins EA, et al.. Comparing reliability and validity of pediatric instruments for measuring health and well-being of children with spastic cerebral palsy. Dev Med Child Neurol. 2002;44:468–476.
16. Oeffinger D, Bagley A, Rogers S, et al.. Outcome tools used for ambulatory children with cerebral palsy: responsiveness and minimum clinically important differences. Dev Med Child Neurol. 2008;50:918–925.
17. Gates PE, Otsuka NY, Sanders JO, et al.. Relationship between parental PODCI
questionnaire and School Function Assessment in measuring performance in children with CP. Dev Med Child Neurol. 2008;50:690–695.
18. Pirpiris M, Graham HK. Uptime in children with cerebral palsy. J Pediatr Orthop. 2004;24:521–528.
19. Cremeens J, Eiser C, Blades M. Factors influencing agreement between child self-report and parent proxy-reports on the Pediatric Quality of Life Inventory 4.0 (PedsQL) generic core scales. Health Qual Life Outcomes. 2006;4:58–65.
20. Gates P, Otsuka N, Sanders J, et al.. Functioning and health-related quality of life of adolescents with cerebral palsy: self versus parent perspectives. Dev Med Child Neurol. 2010;52:843–849.
21. Oeffinger DJ, Rogers SP, Bagley A, et al.. Clinical applications of outcome tools in ambulatory children with cerebral palsy. Phys Med Rehabil Clin N Am. 2009;20:549–565.
22. Vitale MG, Roye EA, Choe JC, et al.. Assessment of health status in patients with cerebral palsy: what is the role of quality-of-life measures? J Pediatr Orthop. 2005;25:792–797.
23. Oladeji OO, Johnston TE, Smith BT, et al.. Quality of life in children with spinal cord injury. Pediatr Phys Ther. 2007;19:296–300.
24. Pirpiris M, Gates PE, McCarthy JJ, et al.. Function and well-being in ambulatory children with cerebral palsy. J Pediatr Orthop. 2006;26:119–124.
25. Hoenig. JM, Heisey DM. The abuse of power: the pervasive fallacy of power calculations for data analysis. Am Stat. 2001;55:1–6.
26. Berger W, Altenmueller E, Dietz V. Normal and impaired development of children’s gail. Hum Neurobiol. 1984;3:163–170.
27. Hillman SJ, Stansfield BW, Richardson AM, et al.. Development of temporal and distance parameters of gait in normal children. Gait Posture. 2009;29:81–85.
28. Johnson DC, Damiano MD, Diane L, et al.. The evolution of gait in childhood and adolescent cerebral palsy. J Pediatr Orthop. 1997;17:392–396.
29. Bell K, Õunpuu S, DeLuca P, et al.. Natural progression of gait in children with cerebral palsy. J Pediatr Orthop. 2002;22:677–682.
30. Aran A, Shalev RS, Biran G, et al.. Parenting style impacts on quality of life in children with cerebral palsy. J Pediatr. 2007;151:56–6060 e51.
31. Feinstein AR. The pre-therapeutic classification of co-morbidity in chronic disease. J Chronic Dis. 1970;23:455–468.
32. McMulkin ML, Baird GO, Gordon AB, et al.. The pediatric outcomes data collection instrument
detects improvements for children with ambulatory cerebral palsy after orthopaedic intervention. J Pediatr Orthop. 2007;27:1–6.
33. McDonald CM, McDonald DA, Bagley A, et al.. Relationship between clinical outcomemeasures and parent proxy reports of health-related quality of life in ambulatory children with Duchenne muscular dystrophy. J Child Neurol. 2010;25:1130–1144.
34. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care. 1989;27supplS217–S232.