Exclusion criteria
1. Studies reporting quality of life for primary sclerosing cholangitis, autoimmune hepatitis.
2. Case reports, conference abstracts, case series, and dissertations were excluded.
3.Using fatigue scales or pruritus scales to assess patients’ quality of life.
4.Number of included cases<=10.
2.4 Data extraction
Two reviewers (X.Y. and J.-M.X.) assessed the full-text publications and extracted data using standardized electronic forms independently. Disagreement between the reviewers about eligibility was resolved through discussion. Data extracted from the selected studies including: author, year of publication, study design, aim of the study, study characteristics, intervention, HRQOL tools, results. For translated questionnaires, we evaluated floor and ceiling effects, missing data, internal consistency and test-retest reliability. Finally, we performed a quality analysis of the included studies following the Hawker’s tool, which provides clear description of ratings, that is, “good,” “fair,” “poor” and “very poor,” and has been designed to assess quality of studies covering a variety of research paradigms[10].
3.Results
3.1 Study selection
The search retrieved 1175 publications and 435 full-text records were obtained after duplicates removed. Studies were published between 1999.1 and 2020.1. 391 studies were excluded with reasons. Finally, 44 studies met the inclusion criteria. A PRISMA flow chart illustrates the studies selection process and reasons for exclusion (Figure 1).
3.2 Overview of included studies
A total of 15 HRQOL questionnaires were used in the included 44 studies, descriptions of the used questionnaires are presented in table 1. The instruments used most frequently were the PBC-40 (n = 21), followed by the SF-36 (n = 19), PBC-27(n=4), CLDQ (n = 3) and NIDDK-QA(n=2), the remaining instruments were uesd only once (Supplementary table 1~3). These instruments were initially designed for various purposes and referred to different concepts. The PBC-40 was the first questionnaire for use in patients with PBC specifically. The SF-36 is a standardized measure of health status, commonly used as a HRQOL questionnaire. The PBC-27 and the PBC-10 were brief vision of PBC-40. The CLDQ, CLDQ-I and LDSI2.0 were liver disease-specific questionnaires. The other questionnaires were developed to measure generic quality of life.
Validity of translated questionnaires
Tweenty-six studies used a translated HRQOL questionnaire and only six reported or referenced a validation of the translated questionnaire[11-16].
Floor/ceiling effects and missing data
Floor and ceiling effects were reported for only two HRQOL questionnaires (PBC-40 and PBC-10).There were no significant ceiling effects but a moderate floor effect was found for the itch domain in PBC-40. The PBC-10 did not have a significant ceiling effect but flooring effects were observed in all 10 items similarly[15, 17].
Internal consistency
Cronbach’s α coefficients for the HRQOL questionnaires ranged from adequate to excellent. The PBC-40 had the highest internal consistency (Cronbach’s α coefficient 0.72–0.95)[17]. PBC-27 Cronbach’s α coefficients ranged from 0.45 to 0.93[13], NIDDK-QA from 0.87 to 0.94 and for PBC-10 it was 0.90[15, 16]. The internal consistency of other questionnaires was not reported.
Test-retest reliability
Test-retest reliability data were available for PBC-40, NIDDK-QA and PBC-10. The test-retest reliability of NIDDK-QA was measured by Pearson Correlation, it was ranged from 0.82 to 0.99. Test-retest reliability was calculated using intraclass correlation coefficients (ICCs) of PBC-40 and PBC-10. The PBC-40 ICCs ranged from 0.83 to 0.96 and for PBC-10 it was 0.945.