Cerebral palsy (CP) is the most common physical disability in childhood (SCPE, 2000). This nonprogressive condition impacts on a child’s movement, balance, sensation, perception, cognition, communication and behaviour (Rosenbaum et al., 2007), and impacts on the child’s Quality of Life (QOL) (Vargus-Adams, 2005; Varni et al., 2005). Epilepsy and secondary musculoskeletal problems are also commonly found (Rosenbaum et al., 2007). QOL and Health-Related Quality of Life (HRQOL) are increasingly being considered as therapeutic goals (Aran, 2010). QOL is defined by the WHO QOL assessment group as an ‘individual’s perception of their position in life in context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns’ (WHOQOL, 1995). Although HRQOL is considered as a subset of QOL, these two are differently constructed and are not interchangeable (Davis and Waters, 2010). HRQOL instruments are broadly classified as generic and disease-specific measures (Waters et al., 2009). Generic instruments allow comparison among children with various conditions; thus, questions are broadly based rather than specific to a particular condition (Waters et al., 2009). Condition-specific instruments may be more likely to detect change with intervention or over time (Aran, 2010).
Valid and reliable generic HRQOL instruments have been developed for children (McCarthy et al., 2002; Varni et al., 2006; Gates et al., 2008; McCullough and Parkes, 2008; Davis et al., 2010). The Pediatric Quality of Life Inventory 4.0 (PedsQL 4.0) generic instrument is particularly useful, being reliable and valid (Varni et al., 2001), widely used and translated into many languages (Amiri et al., 2010; Yang et al., 2010; http://www.pedsql.org) including Thai. Although the use of the Thai PedsQL 4.0 Generic Core Scales was reported in Thai children with chronic diseases (Pongwilairat et al., 2005; Punpanich et al., 2010), no psychometric properties have been reported. It would be expected that the validity of the tool is retained when translated.
A number of valid and reliable HRQOL instruments have been designed specifically for children with CP (Baars et al., 2005; Narayanan et al., 2006; Varni et al., 2006; Waters et al., 2007; Davis et al., 2010). These condition-specific instruments have been reported to measure components in the domains of body functions and structures, activity and participation as well as contextual factors of the International Classification of Functioning Disability and Health (Schiariti et al., 2011). The PedsQL 3.0 CP Module (Varni et al., 2006) was translated into nine languages (http://www.pedsql.org/translations.html) and most recently into Thai by Tantilipikorn et al. (2012), who reported excellent psychometric properties of the translated instrument. It is unclear how differently a generic tool would evaluate HRQOL compared with a condition-specific instrument in the same population with CP. Therefore, this study aims, first, to establish the psychometric properties of the Thai PedsQL 4.0 Generic Core Scales, and second, to compare the findings between this and the Thai PedsQL 3.0 CP Module.
Ethical approval was granted from the Medical Research Ethics Committee of the University of Queensland (Approval No. 2010000917) and two local venues in Thailand: the Queen Sirikit National Institute of Child Health (QSNICH) and the Srisangwan school (Nonthaburi).
Convenience sampling techniques were used for recruiting patients from QSNICH and Srisangwan school. In this paper, ‘parent’ is used for ‘parent or caregiver’. Local physiotherapists recruited children and/or parents who fulfilled all the inclusion criteria. These were parents of children and/or children younger than 18 years of age with no intramuscular injection or muscle lengthening/surgery in the previous 3 months, and no plan for these interventions before retesting. Younger children (2–4 and 5–7years) were recruited from QSNICH, whereas the older children (5–7, 8–12 and 13–18 years) were randomly sampled from Srisangwan school. Sample size was set at 15 per group to provide 80% power at α less than or equal to 0.05 for reliability testing (Donner and Eliasziw, 1987).
Thai PedsQL 4.0 Generic Core Scales
All versions of the PedsQL 4.0 Generic Core Scales were divided into four age groups (2–4; 5–7; 8–12; 13–18 years), with child self-administered (≥5 years) and parent-proxy forms (all ages). Ratings use a five-point scale asking about how much of a problem each item has been during the past 1 month (0=never a problem, 1=almost never a problem, 2=sometimes a problem, 3=often a problem, 4=almost always a problem) for all age groups and versions except the 5–7 year group (child report). This form uses a three-point Likert scale (0=not at all a problem, 2=sometimes a problem, 4=a lot of a problem). Responses are converted into a score from 0 to 100 with a reverse pattern (0=100, 1=75, 2=50, 3=25, 4=0), with a higher score representing better HRQOL (Varni et al., 2001).
Scales include (1) Physical Functioning: eight items, (2) Emotional Functioning: five items, (3) Social Functioning: five items (4) School Functioning: five items. The Psychosocial Health score was obtained by dividing the sum of the mean items of Emotional, Social and School Functioning scales by the number of these items that were answered. The PedsQL 4.0 Generic Core Scales (Thai version) were provided by the Mapi research institute for use in this study (http://www.pedsql.org).
Thai PedsQL 3.0 CP Module
The age groups and response options are similar to the PedsQL 4.0 Generic Core Scales. The scales of the PedsQL 3.0 CP Module include (1) Daily Activities: nine items, (2) School Activities: four items, (3) Movement and Balance: five items, (4) Pain and Hurt: four items, (5) Fatigue: four items, (6) Eating Activities: five items, (7) Speech and Communication: four items. The calculation for the total score is the same as that for the Generic Core Scales. Good to excellent psychometric properties were reported (Tantilipikorn et al., 2012). Data from this version are provided here for comparison with the values of the Thai PedsQL 4.0 Generic Core Scales, using the same participants.
As part of a larger series of studies, the psychometric properties of the newly translated Thai CP Module have been reported elsewhere (Tantilipikorn et al., 2012), but data were concurrently collected for both this and the generic Thai version. A summary of the collection process is provided here.
After recruitment, the researcher reconfirmed that the criteria were fulfilled, consent forms were signed and parents and children participated in the testing of both tools. A general information questionnaire was completed, and then the generic was completed, followed by the condition-specific instrument, as recommended by PedsQL guidelines (http://www.pedsql.org). Both the Thai PedsQL 4.0 Generic Core Scales and the Thai PedsQL 3.0 CP Module were reassessed within 2–4 weeks for test–retest reliability. The parents who could not visit the local venues at this time were interviewed by phone.
Feasibility was determined by the percentage of missing values (Varni et al., 2001), the floor and ceiling effect by calculating the percentage of cases reporting a ‘0’ score (the lowest) and ‘100’ (the highest), respectively (McHorney et al., 1994). Statistical package for social sciences (SPSS Inc., Chicago, Illinois, USA) version 19 was used for analysis. Test–retest reliability was evaluated using the two-way mixed model of intraclass correlation coefficient (ICC3,1). ICC was graded as less than 0.40: poor–fair, 0.41–0.60: moderate, 0.61–0.80: good, 0.81–1.00: excellent (Varni et al., 2005).
Cronbach’s α coefficient evaluated the internal consistency reliability of the group scores (Portney and Watkins, 2009), with α at least 0.70 considered acceptable (Hair, 2006). All significant levels were applied to the group and not at the individual level. Correlation among scales was assessed by Pearson’s product–moment correlation, interpreted as 0–0.25: little or no relationship, 0.25–0.50: fair, 0.50–0.75: moderate–good, more than 0.75: good–excellent (Portney and Watkins, 2009).
The results of the Thai PedsQL 4.0 Generic Core Scales are presented below with data from a concurrent study of the Thai PedsQL 3.0 CP Module (Tantilipikorn et al., 2012) provided for comparison.
A maximum number of 97 parent-proxy reports and 54 child self-reports were available.
Most respondents’ highest education level was below undergraduate (82.5%). Respondents (n=97) were parents 79.4% (mother 87.01%) and other caregivers 20.6%. Five respondents did not complete the retest. The mean age of the parent-proxy report respondents was 39.27±10.38 years (range 19–65 years).
Children of respondents (n=54) were boys (56.7%) with an overall mean age of 11.54±3.41 years (range 5.75–18 years). Type of CP for the child respondents was not identified on the charts in five children (9.3%). When specified, 32 children were diagnosed with diplegia (65.30%), six with athetosis (12.24%), five with quadriplegia (10.20%) and six as ‘other’ (12.24%).
Descriptive data and floor/ceiling effects
The means, SDs, and the floor and ceiling effects for child and parent forms of Thai PedsQL 4.0 Generic Core Scales are presented in Table 1. The Daily Activities parent-proxy report score showed a greater floor effect than the other scales, whereas the Speech and Communication child self-report showed a greater ceiling effect. Children aged 2–4 years only had the parent-proxy report available. Overall, the parent mean scores were lower than the child mean scores (Table 1), and this pattern was maintained when only child self-report (≥5 years) was compared with the parent reports.
The only child self-report missing response was in Emotional Functioning and School Functioning scales at 1.81%. For the parent-proxy report, the percentage of missing item responses was 3.13% for all scales, except School Functioning. Percentage with 80.77% (2–4 years) missing as this scale was not applicable for this age group. Children aged 5–18 years had 11.43% missing data here.
Reliability is reported in two ways: (a) overall reliability of parent-proxy report and self-report with all ages combined as shown in Table 2 and (b) overall reliability for each age group of parent-proxy report (Table 3) and self-report (Table 4).
Overall, test–retest reliability for parent-proxy report (ICC=0.866) and child self-report (ICC=0.870) was excellent. When reliability was examined for each scale, parent report for each scale ranged from 0.625 to 0.866 and child self-report ranged from 0.633 to 0.870.
When reliability was considered for each age group for each scale, generally good–excellent reliability was found for both parent-proxy and child self-report as shown in Tables 3 and 4.
Internal consistency was acceptable (α≥0.7) for the overall parent and child group scores (Table 5), whereas the Emotional, Social and School Functioning scale scores were not acceptable.
Correlation between and within the Thai PedsQL 4.0 Generic Core Scale and the PedsQL 3.0 CP Module
Correlations of the total score between these instruments are 0.717 for parent-proxy report and 0.467 for self-report. The correlations between these instruments for each scale varied as shown in Table 6. Moderate–good correlation was found between Physical Functioning (PedsQL 4.0 Generic Core Scales) and the total score of the PedsQL 3.0 CP Module for both parent-report and self-report.
Within PedsQL 4.0 for parent-report, moderate–excellent correlations were found between the total score and Physical Functioning, Psychosocial Health. Within the PedsQL 3.0 CP Module, moderate–excellent correlations were also found among all scales compared with the total score, except the Pain and Hurt, which was fair.
This study has established the psychometric properties of the Thai PedsQL 4.0 Generic Core Scales as acceptable, and showed that the Thai PedsQL 3.0 CP Module provided better psychometric values for the same group of Thai children with CP, as well as more detailed information on the effects of the CP condition itself.
The Thai PedsQL 4.0 Generic Core Scale
This study has shown acceptable test–retest reliability for the total score, which was similar for parent and child forms across all ages, with the lowest reliability (good range) found in School Functioning. There was some variation across age groups. These findings are similar to a preliminary study of the Chinese version for parents and children with disability aged 2–7 years (Chan et al., 2005). However, the Swedish version studied in healthy school-age children (8–14 years) showed higher reliability than the present study in both parent-report and child-report in all scales (ICC=0.82–0.93) (Petersen et al., 2009).
Mainly, there was good–excellent reliability, with mean scores in each scale similar to those in the original version (Varni et al., 2006). When all the age groups were included, and when each was considered separately, reliability was established for each child and parent scale at least at a ‘good’ level, with School Functioning the poorest for both child and parent. The possible impact of a likely cognitive impairment in these children on their self-report accuracy is unclear. Thus, a parent-proxy evaluation may be more useful if intellectual impairment is suspected.
Little floor effect of the Thai PedsQL 4.0 Generic Core Scales in parent report was found in this study, but was not found at all in any scale of other versions (Reinfjell et al., 2006; Gkoltsiou et al., 2008; Amiri et al., 2010). Conversely, no ceiling effect was found in self-report and only a small percentage was found in some scales of parent-proxy in this Thai version, although the ceiling effect was shown in other versions and study groups (Reinfjell et al., 2006; Gkoltsiou et al., 2008; Amiri et al., 2010). This might be related to the different conditions and age ranges used in the other studies. Minimal missing data showed feasibility with irrelevant information on School Functioning: 2–4 years as this is not appropriate for these young children. Overall, internal consistency was clearly found, but was not achieved for some individual scales, in agreement with the findings from the original English version (Varni et al., 2006); Waters et al. (2009) suggested that internal consistency is sensitive to the number of items, with a higher value reflective of more items even if the items are heterogeneous. A small number of items in some scales here might have contributed to these findings.
In addition, the mean total parent-proxy report (52.71) and child self-report (62.81) scores found in this study are below the cut-off points for at-risk status for impaired HRQOL reported (parent-proxy report=69.4; self-report=69.7) (Varni et al., 2003). Therefore, even using the generic instrument, impaired HRQOL can be confirmed in this group of Thai children with CP.
The Thai PedsQL 4.0 Generic Core Scale and the PedsQL 3.0 CP Module: relationships and differences
A main difference between these two versions is the tailoring of items to the CP condition in the CP Module, which we expected would produce more reliable answers, which also pertained specifically to these individuals. Scales are differently named and contain different items in each of these instrument, making a direct comparison difficult.
Relationships: when considering each scale of these tools, moderate–good correlations were found for parent-report and child self-report among Physical Functioning of Generic Core Scales and Daily Activities, Movement and Balance and School Activities of the CP Module. This is likely because of some similarity in the items in these scales.
The Psychosocial Health scale of the Thai PedsQL 4.0 Generic Core Scales showed fair–good correlation with Pain and Hurt and Fatigue of the Thai PedsQL 3.0 CP Module for both parent-report and self-report, possibly because of some focus on similar aspects. Only those scales with satisfactory internal consistency are included here. Interestingly, in the Thai PedsQL 4.0 Generic Core Scales, three scales were not internally consistent, although in the original version (2006) and other translated versions conducted in other populations (Reinfjell et al., 2006; Amiri et al., 2010), the internal consistency of all scales of this instrument, except for Social Functioning of self-report (Varni et al., 2006), was established. As the same cohort was reliable and consistent using the Thai PedsQL 3.0 CP Module, it seems most likely that the differences found in the consistency of those three scales on the Thai PedsQL 4.0 could be related to internal issues such as the accuracy of the Thai translation. Despite this limitation, the instrument was reliable for the total score, and thus still has utility for this CP population.
In addition, in both versions, the parent and child scores clearly showed that parents rated children lower than children rated themselves across all areas evaluated, in agreement with Majnemer et al., (2008). This response pattern has also been found in children with other chronic conditions (Majnemer et al., 2008), but is reversed for healthy children, where parents often rated their children’s well-being higher than did the children (Eiser and Morse, 2001).
Differences: although the internal consistency of School Functioning in the Thai PedsQL 4.0 Generic Core Scales was unacceptable, the implications of the difference between School Functioning (PedsQL 4.0) and School Activities (PedsQL 3.0) should still be considered. These scales might determine different aspects as the correlation between these is fair for parent-proxy report and there is no correlation for self-report. For example, items of School Functioning (PedsQL 4.0) are related to problems of concentration in a classroom or missing school because of sickness, whereas School Activities (PedsQL 3.0) queried using a pen/pencil or a computer keyboard. Again, the CP Module seems to include information more relevant to the CP population.
Although both versions were feasible, the Thai PedsQL 3.0 CP Module showed higher test–retest reliability for the total scores from both parent and child self-report overall, and in each age group, suggesting that the parents and children were able to respond more consistently when questions were specifically directed to issues relevant to them/their condition. Internal consistency was also higher in the CP Module. The Thai PedsQL 3.0 CP Module may also be more feasible as fewer missing item responses were found here than in the Generic Core Scales. This may be because both parents and children find some of the generic items irrelevant to their lived experiences, implying that the CP Module may better determine HRQOL of children with CP.
The Thai PedsQL 4.0 Generic Core Scales could still be useful for these children as it measures some different elements from the PedsQL 3.0 CP Module, especially psychosocial aspects. For example, the Generic Core Scales contain items about difficulty keeping up with others when playing. However, although different, this may not be very useful information as most children with CP would be expected to experience some difficulty with this, and it is unlikely to improve with either maturation or even intervention. Therapists must decide which instrument to use in children with CP depending on whether the user plans to compare against children with other conditions, or even typically developing children, rather than against others with CP. The findings imply that both instrument could be used for children with CP to assess many aspects of their HRQOL but that more specific details are provided, with better psychometric properties, using the CP Module.
This study showed that the Thai PedsQL 4.0 Generic Core Scales are reliable and feasible, with some limitations in internal consistency of three specific scales in a group of Thai children with CP, and that for the same group, the Thai PedsQL 3.0 CP Module provides an advantage in terms of better psychometric properties and more CP-specific scales and items. Although the total scores of the two instruments correlate, they measure some different elements, with the condition-specific Module providing some advantages to the CP population. Careful and reasoned selection of one tool over the other can now occur, and sometimes, it may be desirable to use both instruments to provide the most complete evaluation of HRQOL in children with CP.
The authors thank all parents/caregivers and children with CP for their kind participation. The authors sincere thanks are due to Dr James W. Varni for his permission to use the tools. The authors sincere thanks are also due to the physiotherapists at QSNICH, especially Thanarat Ratanon and Srisangwan school under foundation for Welfare of Crippled Children, for their cooperation with the recruitment.
Conflicts of interest
There are no conflicts of interest.
Amiri P, Emad MA, Sara J, Farhad H, James V, Fazlollah G, et al. Reliability and validity of the Iranian version of the Pediatric Quality of Life Inventory™ 4.0 Generic Core Scales in adolescents. Qual Life Res. 2010;19:1501–1508
Aran APreedy VR, Watson RR. Quality of life in children with cerebral palsy. Handbook of disease burdens and quality of life measures. 2010 New York Springer:2453–2468
Baars RM, Atherton CI, Koopman HM, Bullinger M, Power M. The European DISABKIDS project: development of seven condition-specific Modules to measure health related quality of life in children and adolescents. Health Qual Life Outcomes. 2005;3 doi: 10.1186/1477-7525-3-70
Chan LF, Chow SM, Lo SK. Preliminary validation of the chinese version of the pediatric quality of life inventory. Int J Rehabil Res. 2005;28:219–227
Davis E, Waters EPreedy VR, Watson RR. Children with cerebral palsy, psychometric analysis and quality of life. Handbook of disease burdens and quality of life measures. 2010 New York Springer:3657–3670
Davis E, Shelly A, Waters E, Davern M. Measuring the quality of life of children with cerebral palsy: comparing the conceptual differences and psychometric properties of three instruments. Dev Med Child Neurol. 2010;52:174–180
Donner A, Eliasziw M. Sample size requirements for reliability studies. Stat Med. 1987;6:441–448
Eiser C, Morse R. Can parents rate their child’s health-related quality of life? Results of a systematic review. Qual Life Res. 2001;10:347–357
Gates PE, Otsuka NY, Sanders JO, McGee-Brown J. Relationship between parental PODCI questionnaire and School Function Assessment in measuring performance in children with CP. Dev Med Child Neurol. 2008;50:690–695
Gkoltsiou K, Dimitrakaki C, Tzavara C, Papaevangelou V, Varni JW, Tountas Y. Measuring health-related quality of life in Greek children: psychometric properties of the Greek version of the Pediatric Quality of Life Inventory(TM) 4.0 Generic Core Scales. Qual Life Res. 2008;17:299–305
Hair JF Multivariate data analysis. 20066th ed. Upper Saddle River, New Jersey Pearson Prentice Hall
Majnemer A, Shevell M, Law M, Poulin C, Rosenbaum P. Reliability in the ratings of quality of life between parents and their children of school age with cerebral palsy. Qual Life Res. 2008;17:1163–1171
McCarthy ML, Silberstein CE, Atkins EA, Harryman SE, Sponseller PD, Hadley-Miller NA. 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
McCullough N, Parkes J. Use of the child health questionnaire in children with cerebral palsy: a systematic review and evaluation of the psychometric properties. J Pediatr Psychol. 2008;33:80–90
McHorney CA, Ware JE Jr., Lu JF, Sherbourne CD. The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups. Med Care. 1994;32:40–66
Narayanan UG, Fehlings D, Weir S, Knights S, Kiran S, Campbell K. Initial development and validation of the Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD). Dev Med Child Neurol. 2006;48:804–812
Petersen S, Hagglof B, Stenlund H, Bergstrom E. Psychometric properties of the Swedish PedsQL, Pediatric Quality of Life Inventory 4.0 Generic Core Scales. Acta Paediatr. 2009;98:1504–1512
Pongwilairat K, Louthrenoo O, Charnsil C, Witoonchart C. Quality of life of children with attention-deficit/hyperactivity disorder. J Med Assoc Thai. 2005;88:1062–1066
Portney LG, Watkins MP Foundations of clinical research: applications to practice. 2009 Upper Saddle River, New Jersey Pearson Prentice Hall
Punpanich W, Boon-Yasidhi V, Chokephaibulkit K, Prasitsuebsai W, Chantbuddhiwet U, Leowsrisook P, et al. Health-related Quality of Life of Thai children with HIV infection: a comparison of the Thai Quality of Life in Children (ThQLC) with the Pediatric Quality of Life Inventory version 4.0 (PedsQL 4.0) Generic Core Scales. Qual Life Res. 2010;19:1509–1516
Reinfjell T, Diseth TH, Veenstra M, Vikan A. Measuring health-related quality of life in young adolescents: reliability and validity in the Norwegian version of the Pediatric Quality of Life Inventory 4.0 (PedsQL) Generic Core Scales. Health Qual Life Outcomes. 2006;4:61
Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M, Damiano D, et al. A report: the definition and classification of cerebral palsy april 2006. Dev Med Child Neurol Suppl. 2007;109:8–14
Schiariti V, Fayed N, Cieza A, Klassen A, O’Donnell M. Content comparison of health-related quality of life measures for cerebral palsy based on the international classification of functioning. Disabil Rehabil. 2011;33:1330–1339
. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Dev Med Child Neurol. 2000;42:816–824
Tantilipikorn P, Watter P, Prasertsukdee S. Feasibility, reliability, and validity of the Thai version of the Pediatric Quality of Life Inventory 3.0 cerebral palsy Module. Qual Life Res. 2012 doi: 10.1007/s11136-012-0161-3. [Epub ahead of print; 25 March 2012]
Vargus-Adams J. Health-related quality of life in childhood cerebral palsy. Arch Phys Med Rehabil. 2005;86:940–945
Varni JW, Seid M, Kurtin PS. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 Generic Core Scales in healthy and patient populations. Med Care. 2001;39:800–812
Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr. 2003;3:329–341
Varni JW, Burwinkle TM, Sherman SA, Hanna K, Berrin SJ, Malcarne VL, Chambers HG. Health-related quality of life of children and adolescents with cerebral palsy: hearing the voices of the children. Dev Med Child Neurol. 2005;47:592–597
Varni JW, Burwinkle TM, Berrin SJ, Sherman SA, Artavia K, Malcarne VL, Chambers HG. The PedsQL in pediatric cerebral palsy: reliability, validity, and sensitivity of the Generic Core Scales and Cerebral Palsy Module. Dev Med Child Neurol. 2006;48:442–449
Waters E, Davis E, Mackinnon A, Boyd R, Graham HK, Kai Lo S, et al. Psychometric properties of the quality of life questionnaire for children with CP. Dev Med Child Neurol. 2007;49:49–55
Waters E, Davis E, Ronen GM, Rosenbaum P, Livingston M, Saigal S. Quality of life instruments for children and adolescents with neurodisabilities: how to choose the appropriate instrument. Dev Med Child Neurol. 2009;51:660–669
. The World Health Organization Quality of Life assessment (WHOQOL): position paper from the World Health Organization. Soc Sci Med. 1995;41:1403–1409
Yang X, Xiao N, Yan J. The PedsQL in pediatric cerebral palsy: reliability and validity of the Chinese version pediatric quality of life inventory 4.0 Generic Core Scales and 3.0 cerebral palsy Module. Qual Life Res. 2011;20:243–252
© 2013 Lippincott Williams & Wilkins, Inc.