38 RESULTS Figure 2.1 shows the PRISMA diagram of our systematic review. The final search yielded 860 articles, written between 1987 and 2019. 48 studies were eligible for inclusion. Main reasons for exclusion were: 1) used educational level as measurement for HL, and 2) study type. STUDY CHARACTERISTICS Figure 2.2 gives an overview of the main characteristics of the included studies. We identified 38 cross-sectional, cohort or mixed-method studies, four qualitative, and six intervention studies, all in the English language. Most studies had sample sizes below 200 (n=33), were conducted in the United States (n=35) and focused mainly on dialysis and transplant patients (n=38). Only seven studies measured multiple HL domains, instead of just functional HL. Details on authors, year of publication, study population, sample size, and used HL screener are in Table 2.1, 2.2 and 2.3. QUALITY ASSESSMENT Nine quantitative studies and one qualitative study were of high quality. Nine quantitative studies, two qualitative studies and one intervention study were of moderate quality. The other 26 studies were of low quality. The risk for external validity bias was high: only 2 studies could fully ascertain the study population was a good representation of the total population. In 25 studies, participation bias was a risk: sample sizes were often not justified or participation rates were low. Within the domains reporting and internal validity, two criteria commonly caused risks of bias: 1) limited adjustment for confounders, 2) not reporting actual probability values (e.g. 0.035 rather than <0.05). In qualitative studies, bias risks were often a consequence of inappropriate methodology: studies did for example not justify sampling procedure and data saturation. Most intervention studies used weak non-randomized control study designs, which led to low quality ratings. Tables 2.1-2.3 show the overall quality rating and Supplemental Tables 3b-3d provide details on the domain ratings for each study.
RkJQdWJsaXNoZXIy MTk4NDMw