Table 4 Evidence grading and overall ratings for the relevance, comprehensiveness, and comprehensibility of the included patient-reported outcome measures (PROMs) for health-related quality of life (HRQOL) assessment in children with cancer
PROM | Relevance | Comprehensiveness | Comprehensibility | ||||||
|---|---|---|---|---|---|---|---|---|---|
Rating of Results | Justification for rating of results | Quality of evidence | Rating of results | Justification for rating of results | Quality of evidence | Rating of results | Justification for rating of results | Quality of evidence | |
DISABKIDS 12/37 | + | No cancer patients involved, but all items reflect concepts of the conceptual model for QOL provided by Anthony et al. [6]. | Low | − | Mapping indicates underrepresentation of physical and cognitive issues (DCGM-12: 0 physical, 0 cognitive items DCGM-37: 3 physical, 1 cognitive items). | Low | − | Items, response options, and recall period are age-appropriate; but unclear if/how difficulties identified in studies were addressed [55, 69]. | Low |
KIDSCREEN 10/27/52 | + | More than 90% of the items (except items on financial situation: KIDSCREEN 27: 2 items, KIDSCREEN 52: 3 items) match concepts of the QOL model by Anthony et al. [6]. | Very low | − | Mapping indicates underrepresentation of physical and cognitive issues. KIDSCREEN 10, 27, 52 cover 0, 2, 2 physical and 1, 1, 1 cognitive items, respectively. Core symptoms like pain or nausea are missing | Very low | + | Studies state that no difficulties [48, 73]. This impression is shared by the reviewers. | Very low |
Kid KINDL-R Generic | + | When asked about the relevance of the whole questionnaire, 80% report perceiving it as very relevant or relevant [76]. All items could be mapped onto the model of QOL by Anthony et al. [6]. | Low | − | Mapping indicates underrepresentation of physical (3 items) and cognitive (2 items) issues in the generic tool, which might explain why 20% of chronically ill children did not perceive the questionnaire as a whole as relevant [76]. | Low | + | “95.5% of the chronically ill children found the questions easily understandable” [76]. | Low |
Kid-Kiddo KINDL-R Oncology | + | All items could be mapped onto the QOL model by Anthony et al. [6]. | Very low | − | Mapping indicates that the tool is not comprehensive. Cognitive issues are not assessed at all. From the psychological domain, only treatment burden (“bothered”; 12 items) and moodiness (2 items) are assessed. Neither other forms emotional distress nor positive mental health aspects are covered. | Very low | − | The reviewers rated the comprehensibility as insufficient due to very complex response-options. Many items require more than one response: For symptoms, children must indicate frequency and the resulting burden. For treatment- or procedure-related issues, a conditional item is followed by a frequency and a burden rating. | Very low |
PAC-QoL Child | + | Items mapped well onto the conceptual model by Anthony et al. [6], except four items covering involvement in care, which was added as a new subdomain in this review. | Very low | + | All domains and nearly all subdomains (except self-esteem and treatment-burden) are covered. | Very low | − | Results of studies available at the time indicate insufficient comprehensibility [62]. | Very low |
PedsQL Generic Core Scale 4.0 | + | All items cover issues reflecting the QOL model by Anthony et al. [6]. | Very low | − | No items assess positive psychological functioning, self-esteem, or body image. | Very low | + | Reviewers' ratings indicate sufficient comprehensibility because instructions, items, and response-options are considered age-appropriate. | Very low |
PedsQL Cancer Module 3.0 | + | All items cover issues reflecting the QOL model by Anthony et al. [6]. | Very low | − | No items on positive psychological functioning are assessed. The social domain does only contain items on involvement in care, whereas the social context of family, peers, school, and leisure time is missing. | Very low | + | Reviewers' ratings indicate sufficient comprehensibility because instructions, items, and response-options are considered age-appropriate. | Very low |
PedsQL Brain Tumor Module | + | All items cover issues reflecting the QOL model by Anthony et al. [6]. | Very low1 | − | No items on positive psychological functioning or social domain. | Very low1 | + | Reviewers' ratings indicate sufficient comprehensibility because instructions, items, and response-options are considered age-appropriate. | Very low1 |
PROMIS Ped Profile 25/37/49 | + | All items cover issues reflecting the QOL model by Anthony et al. [6]. Jones et al. [105] did not report any general relevance issues when comparing child- and parent ratings on relevance. | Moderate | ± | Healthy children and pediatric asthma patients had nothing to add [85]. However, this might not be representative for children with cancer. (Children with chronic kidney or Crohn’s disease and adult cancer patients had disease-specific issues to add; see [106, 107]). Mapping indicates that cognitive as well as positive psychological issues are missing. Even though the physical domain takes approximately the half of items, none covers nausea, a core symptom in oncology. | Moderate | + | PROMIS is the only inventory reporting comprehensibility for items, response-options, and instructions separately and results indicate very good comprehensibility [85]. | Moderate |
QOLCC-7-12 | + | Yeh et al. [64] report that besides quantitative criteria, “items were retained if at least two pediatric oncology specialists or three patients from the qualitative study considered them clinically important” (p.163). Most items mapped well onto the model by Anthony et al. [6], except three items on knowledge about cancer and three items on involvement in care—aspects which were added as new subdomains. | Very low | + | Mapping indicates that all relevant domains are represented, even though the physical domain takes a smaller proportion than in other cancer-specific tools. | Very low | + | Important study (Huang and Yeh, 2000 cited in [100]) not accessible due to language barriers (request for English abstract unanswered). Reviewers' ratings indicate sufficient comprehensibility, because item wording was considered age-appropriate. No information about recall-period or response-options available. | Very low |
SQOLPOP | ? | In case of ‘very low’ evidence, the rating of results should rely on reviewers’ ratings only. However, as no review copy or item list was available, it was impossible to make any ratings. | Very low | ? | In case of ‘very low’ evidence, the rating of results should rely on reviewers’ ratings only. However, as no review copy or item list was available, it was impossible to make any ratings. | Very low | ? | In case of ‘very low’ evidence, the rating of results should rely on reviewers’ ratings only. However, as no review copy or item list was available, it was impossible to make any ratings. | Very low |
TACQOL | + | All items cover issues reflecting the QOL model by Anthony et al. [6]. | Very low | + | All domains of the QOL model by Anthony et al. [6] are covered and take a reasonable proportion. | Very low | + | Following the impression of Vogels et al. [58] that the questionnaire is “self-explanatory” (p. 460), comprehensibility was rated sufficient. | Very low |