Key points are not available for this paper at this time.
Patients who have end-stage renal failure and are treated by hemodialysis (HD) face a stressful chronic illness with a demanding treatment regimen that affects quality of life. Quality-of-life domains can be measured by assessment questionnaires that are easy to complete, reliable, valid, and sensitive to change. There is current interest in HD regimens that provide more frequent treatments (e.g., daily) than the conventional thrice weekly. Improvement in quality of life by these regimens has been reported. A published prospective, cohort, controlled study (London Daily/Nocturnal Hemodialysis Study) included the results of a number of quality-of-life indicators that were applied to the study patients. In general, the indicators used were well established and of proven validity. Included was one single question that was added intuitively and had not received previous validation: "How long does it take you to recover from a dialysis session?" The responses to this question allow the validation of this simple question as a tool to be used in HD clinical research. Twenty-three patients who were treated by frequent HD (5 to 7 d or nights) and 22 control subjects who were treated by thrice-weekly dialysis were studied during an 18-mo period. The "time to recovery" question was administered along with a battery of renal disease-specific questionnaires and the Generic Medical Outcomes Survey 36 Item-Short Form (SF-36) plus the global Health Utilities Index. Missing data rates, reliability over time, construct validity, and sensitivity to change were assessed from the "time to recovery" responses by standard methods. The question was administered on a total of 314 occasions and answered successfully on 313. The test-retest correlation over 3-mo intervals was highly significant (r = 0.962, P = 0.000; n = 100). Convergent construct validity was established by significant correlations between time to recovery and fatigue (r = 0.38, P = 0.000; n = 313), dialysis stress (r = 0.348, P = 0.000), disease stress (r = 0.374, P = 0.000), SF-36 subscales especially vitality (r = -0.356 P = 0.000), and the Health Utilities Index (r = -0.232, P = 0.000). These scales captured mainly physical or physiologic domains. Divergent construct validity was established by lack of correlations between "time to recovery" and a number of subscales that captured mainly emotional or psychosocial domains, e.g., SF-36 subscale for "role emotional" (r = -0.102, NS) and dialysis stressors such as access problems (r = -0.015, NS) or equipment malfunction (r = 0.032, NS). Test sensitivity was established when the conventionally dialyzed group showed no significant difference in time to recovery between baseline and other time periods, whereas the daily/nocturnal group had a significant reduction between baseline (while on conventional dialysis) and the result at each other time period (minimum P = 0.05). There also was a significant difference between the control and experimental groups over time (ANOVA P = 0.000). The response to the question, "How long does it take you to recover from a dialysis session?" is interpreted easily, is easy to which to respond, shows stability over time by test-retest, shows both convergent and divergent validity, and is sensitive to change. As such, it should be considered as a standard question in HD-related studies in which a health-related quality-of-life outcome is examined.
Lindsay et al. (Fri,) studied this question.