Background

Hunter syndrome, mucopolysaccharidosis II (MPS II), is a rare, X-linked, progressive, multi-systemic, lysosomal storage disorder caused by a deficiency of the enzyme iduronate-2-sulfatase [1] with an estimated incidence of 1 per 170,000 male births [2]. A wide spectrum of clinical disease occurs from attenuated to severe, with highly variable rates of progression and degree of organ involvement, resulting in significant impairment of patient’s function and quality of life [3].

In order to monitor disease progression and evaluate treatment in clinical trials, a MPS II-specific instrument, the Hunter Syndrome-Functional Outcomes for Clinical Understanding Scale (HS-FOCUS) Questionnaire, was developed by a group of experts led by Dr J Muenzer. Parent-completed and patient self-reported Questionnaires were created through the literature review and input from expert clinicians, parents and patients. Assessment of measurement properties of the HS-FOCUS was recently undertaken [4] and revealed that it was a valid and reliable instrument but could be further improved by reducing items that were redundant.

This communication is a follow-up to the previous publication [4] with the objective to reduce the number of items of the HS-FOCUS in order to have a shorter and more efficient instrument that also decreases the respondent burden.

Methods

Data collected in a Phase II/III trial (NCT00069641) [5] as used in the validation study [4] were further analysed to reduce the items and re-assess the measurement properties of the shortened HS-FOCUS.

The original HS-FOCUS could be completed by parents/caregivers (68 items) or patients themselves (54 items). In both, items were grouped into six function domains: Walking/standing, Reach/grip, Sleeping, School/work, Activities and Breathing; and a satisfaction-with-function and a botheredness-with-function domains [4]. The item response scale ranged from 0 being ability to complete the activity ‘without ANY difficulty’ to 4 as ‘UNABLE to do so’. Average domain scores were computed using item responses if less than half of them were missing or ‘not applicable’.

Parents of patients of all ages responded to the parent-completed HS-FOCUS and patients aged ≥12 years to the self-reported version. The same participants also completed the following measures:

  • The Childhood Health Assessment Questionnaire (CHAQ) A 30-item disease-specific instrument [6] that comprises eight domains: dressing, arising, eating, walking, reach, grip, hygiene and activities. Each domain is scored from 0 ‘without ANY difficulty’ to 3 ‘UNABLE to do’, and the average of the domain makes up the disability index score (DIS).

  • Health Utilities Index (HUI3) A family of generic health profiles and preference-based systems with eight attributes: vision, hearing, speech, ambulation, dexterity, emotion, cognition and pain, with five to six levels per attribute [7]. It includes self-assessed and proxy-assessed forms. The scoring system provides utility (preference) scores on a scale where dead = 0 and perfect health = 1.

  • Childhood Health Questionnaire (CHQ) A generic Questionnaire that measures 12 unique physical and psychosocial concepts in the children’s version CHQ-CF87 and 14 in the parent-completed CHQ-PF50: physical functioning, role/social–emotional and role/social–behavioural, role/social–physical, bodily pain, general behaviour, mental health, self-esteem, general health perceptions, change in health, parental impact–emotional, parental impact–time, family activities and family cohesion. Concept scores may also be combined to derive overall physical and psychosocial scores [8] from 0 to 100.

Statistical analysis of the parent-completed and patient self-reported responses were analysed separately using Stata/MP Ver.11.0 [9].

An iterative process was used to identify potentially redundant or poorly performing items using distributional characteristics of the HS-FOCUS item responses (mean, standard deviation, % missing and % at floor and ceiling). Each item identified was further reviewed and discussed individually and in the domain context. Input from two paediatric clinicians with long standing experience in treating MPS II was sought throughout the process.

Potentially redundant items were identified by any of the following criteria: high (≥60 %) percentage of missing responses suggesting the item was less relevant to MPS II or it was not clear; floor effects with ≥60 % responding ‘without ANY difficulty’ suggesting the item had a low impact on the majority of MPS II patients; high inter-item correlations (>0.80) suggesting item was redundant; low item-to-total scale correlation (<0.30) and poor clinical relevance. Items were examined by clinical experts for fit within the domain, who also ensured that no clinically relevant items were deleted.

Measurement properties of the shortened HS-FOCUS were then re-assessed. Internal consistency was assessed using Cronbach’s alpha, with α > 0.70 considered acceptable [10–12]. Test–retest reliability was considered acceptable when intra-class correlation coefficient (ICC) >0.70 among stable subjects defined as those whose CHAQ DIS score change did not exceed ±0.13 between baseline and week 18 [13]. Concurrent validity was evaluated with Spearman rank order correlations with CHAQ, CHQ and HUI3 [13]; correlations ≥0.30 between these measures were anticipated. Correlation with the original HS-FOCUS [4] was also calculated to support that content validity was maintained. Known-groups validity was determined using analysis of variance (ANOVA) to compare the mean of the shortened HS-FOCUS domain scores among three levels of disability: mild (CHAQ DIS ≤0.63); mild-moderate (CHAQ DIS >0.63 to ≤1.75) and moderate (CHAQ DIS >1.75) [14]. The extent to which the shortened HS-FOCUS could detect change in patients’ health status was assessed by comparing the mean domain scores in patients who showed improvement measured by a CHAQ DIS decrease ≥0.13 [14] versus those who did not.

Results

Altogether, 49 MPS II patients 12 years and older and 84 parents completed the HS-FOCUS. The average age when diagnosed was 5 years (SD 4.75; median 4; P10–P90: 0–13). Patients’ demographic and clinical characteristics have been previously described in the HS-FOCUS validation study [4].

Table 1 shows the number of items retained for each HS-FOCUS domains and the reasons for item removal (the full content of the items is not shown because HS-FOCUS is proprietary). One of the items in the Sleeping domain was moved to Breathing, and two items in the Activities domain were moved to Grip/reach. Subsequently, the Sleeping domain were removed from the HS-FOCUS because after moving one item to breathing the remaining items were considered inadequate, and also because there are other well-established sleep measures (e.g., Tayside Children’s Sleep Questionnaire [15] or the Infant Sleep Questionnaire (ISQ) [16] ) that could be used to supplement the HS-FOCUS.

Table 1 Item reduction in the patient and parent versions of the HS-FOCUS
Full size table

Consequently, both patient-completed (18 items) and parent-completed (21 items) Questionnaires of the shortened HS-FOCUS each contained five functional domains ranging from 0 to 3 (authors’ note: The HS-FOCUS is currently not publicly available; plan for sharing is under consideration, please contact Shire for more information). Table 2 shows the correlation between the shortened version and the original HS-FOCUS was above 0.85 except for the parent version’s Activities score (r = 0.75), supporting that content validity was maintained.

Table 2 Correlation coefficient between the original and shortened HS-FOCUS domains scores at baseline
Full size table

Except for the Breathing domain in the patient’s sample, the internal consistency reliability for all five domains in the shortened HS-FOCUS was very good, with α ranging from 0.67 to 0.90 and 0.76 to 0.87 in the patients and parents’ sample, respectively. Overall, complete data on the HS-FOCUS and the CHAQ at baseline and week 18 was available for 82 parents and 48 patients 12 years old and above, from which 30 and 14 were rated stable, respectively. Good test–retest reliability was found with ICC ranging from 0.72 to 0.97 using patient responses and from 0.64 to 0.90 with parent responses. Concurrent validity indicated domains were consistently correlated to other instruments in the expected way. Moderate (>0.3) to high (>0.6) Spearman correlations were found between all patient self-reported domains and most CHAQ Scales; moderate to high correlations were found with the CHQ physical function, role/social physical, bodily pain and self-esteem scores, but low correlation (<0.3) with behaviour-related items, general health perception, change in health and family cohesion. Moderate to high correlations were found with the HUI3 utility scores: hearing, ambulation, dexterity, pain and overall HRQL utility. A similar correlation structure was found for the parent-completed responses (data not shown).

Excellent known-groups validity was shown with statistically significant differences (P < 0.01) in the ANOVA for most shortened HS-FOCUS domains between groups (Table 3). Results also indicated ability to detect changes in the patient’s condition with a statistically significant difference (P < 0.01) in score changes between the two groups in all five domains (Table 4).

Table 3 Known-groups validity for the shortened HS-FOCUS domains
Full size table
Table 4 Ability of the shortened HS-FOCUS domains to detect change
Full size table

Conclusions

The shortened HS-FOCUS retains the items that assess the most critical functions of patients with MPS II, and has shown to be a reliable, valid and responsive tool. For rare and debilitating diseases such as MPS II, good compliance with any instrument used to monitor disease progression over time and evaluate treatment benefits in clinical trials is essential. At the same time, it is important to reduce the respondent burden, especially in paediatric populations. Respondent fatigue when completing the Questionnaire may affect the validity of the scores [17]; therefore, reducing the length of the Questionnaire is expected to have a positive effect.

The current study is limited by the small sample size overall and by unbalanced groups when assessing known group validity; the latter may cause violation to the homogeneity of variance assumption of ANOVA. In addition, the use of mean imputation has the risk of creating a biased score; scoring should be further studied. Analyses were based on data of the original HS-FOCUS, and the complete validation of the shortened HS-FOCUS will be conducted in the future as collection of more data is ongoing.