Promoting Excellence : ESRD Table of Contents : Statistical Summary of Quality of Life for Dialysis Patients Questionnaire Study Findings


Patients with end-stage renal disease (ESRD) receiving dialysis treatment (n=165) in three different dialysis units (50 patients in Washington, DC; 55 patients in Morgantown, WV; and 60 patients in Mineola, NY) were asked to respond to a series of questions focusing on various dimensions of quality of life (QOL), satisfaction with life, and satisfaction with the care they received from clinical staff while receiving dialysis treatment. Questions asked by interviewers included the 16-item McGill Quality of Life Questionnaire (MQOL) plus the single item overall measure of QOL[i], the Satisfaction with Life Scale[ii] (SWLS), and two patient satisfaction scales, including an adaptation from the Kidney Disease QOL instrument[iii] and the DiMatteo and Hays patient satisfaction instrument.[iv] In addition, a series of questions was administered for the purpose of exploring the impact of other psychosocial dimensions on overall QOL among individuals with kidney disease. Demographic information also was collected for each patient to determine variations in outcomes based on these factors. And finally, medical information was collected, including length of time on dialysis (in months), physical symptoms experienced, presence of pain-related symptoms, the clinician's Karnofsky rating[v], and average level of hemoglobin (in gm/dl), albumin (in gm/dl), and the patient's Kt/V ratio.

Statistical Methods

The analysis strategy had two main objectives. The first objective was to document, in an exploratory analysis, the relationship between the patients' various medical and psychosocial characteristics and the QOL indices. Total scores for each of the QOL and patient satisfaction scales were calculated and used to test associations with other factors. In addition, total scores from each of the four McGill QOL subscales (i.e., physical, psychological, existential, and support domains) were calculated and tested against other factors. Pearson correlations, independent sample t-tests, one-way ANOVAs with a posteriori contrast tests (i.e., least-significant difference), and χ2 tests were used as appropriate, using 2-sided significance levels ≤ .05.

The second objective of the analysis was to explore the development of a new scale that could be used to document the impact of other potential dimensions on QOL, including spiritual beliefs and support networks. Patients were asked to respond to twelve questions on issues focusing on the meaning and value of spiritual beliefs and level of support and comfort in talking about their illness to family or friends, other patients, a social worker or therapist, a priest, rabbi, or other religious person, or their clinician. To explore the potential underlying dimensions associated with the 12-item questionnaire, subjects' responses to the questionnaire were subjected to a principal components analysis (PCA). Bartlett's test of sphericity and the Kaiser-Meyer-Olin statistic were used to further test the strength of the association among variables. Reliability analysis using Cronbach's alpha was also conducted on the reduced scale to test the internal consistency of scale items. Further reliability analysis was conducted using a split-half analysis and interpreted by the Spearman-Brown coefficient.After the final solution was reached with the PCA and reliability analyses, regression-method factor scores were calculated using the factor score coefficients. The resulting subscales were then correlated with the other validated QOL indices, biomarkers, and patient satisfaction scales (described above) to determine the associations among the scales and subscales. Further differences by patient characteristics were also explored using statistical tests, as appropriate.


Although nephrologists have for the most part traditionally relied upon one-dimensional measures of physical functioning (for example, by examining levels of albumin and hemoglobin and Kt/V ratio) to assess QOL among patients with end stage renal disease, the findings clearly indicate the importance of psychosocial factors in defining a patient's overall QOL. The results also indicate interesting associations between measures of QOL and various patient characteristics.Results of the PCA are presented first since these results are incorporated into the subsequent analyses on associations with other measures of QOL.

The PCA and reliability analyses resulted in two constructs: spiritual beliefs and patient support network. These two resulting factors accounted for 45% and 19% of the variance, respectively, for a total of 64%.The final standardized item alpha for the reliability analysis was .84. In addition, the split-half reliability analysis indicated strong correlations between one half of the variable set and the other (correlation = .80) and a strong Spearman-Brown coefficient of .89. The final "spiritual beliefs" construct consisted of four items and the "support network construct" consisted of five items (from the original 12-item questionnaire).

Spiritual beliefs were significantly and positively associated with the following subscales/scales: The MQOL overall score (r(138) = .25, p<.01), the MQOL Existential and Support Subscales (r(150) = .34 and r(148) = .30, respectively, p<.001), the MQOL single item measure (r(157)= .18, p<.05), and the SWLS (r(155) = .21; p<.01). In addition, spiritual beliefs were significantly associated with patients' perceived support from dialysis staff (r(155) = .22, p<.01). Age was significantly and positively associated with spiritual beliefs (r(153) = .21, p<.01), while women compared to men (t(149)=3.57, p<.001) and racial/ethnic minorities compared to whites (t(150)=3.68, p<.001) were likely to have higher scores on the spiritual beliefs construct.

Patients who were more likely to speak about their illness to others in their support network and to gain comfort from this (i.e., support network construct) also were significantly more likely to score higher on the following QOL measures compared to patients that scored lower on the support network measure: The MQOL Existential subscale (r(150)= .21, p=.01) and the MQOL Support subscale (r(151)= .29, p<.001).Patients scoring higher on the support network construct also were more likely to be satisfied with their doctors and staff (t(136)=2.61, p=.01).

Whites compared to racial/ethnic minorities were likely to score significantly lower on the support network subscale (t(97)=2.28, p<.05), while minorities scored significantly lower on the MQOL Psychological subscale compared to whites (t(85)=2.28, p<.05). In addition, women were significantly more likely than men to score higher on the McGill Support subscale (t(136)=2.61, p=.01), while minorities were significantly more likely to report greater overall average satisfaction with doctors and staff compared to whites (t(157)=2.50, p<.05).

The relationships between the number of troublesome symptoms reported, presence of pain-related symptoms reported by patients and QOL scores were also explored. Fifty-five percent of patients reported having no troublesome symptoms, while 28% and 17% of patients reported having one, or more than one, troublesome symptom, respectively. There was a significant inverse dose-response relationship between the MQOL total score (F(2, 138) = 38.8, p<.0001), the MQOL physical subscale (F(2, 158) = 180.3, p<.0001), the single item global measure of QOL (F(2, 162) = 12.5, p<.0001) and the number of symptoms reported by patients. The total score on the SWLS also was significantly and inversely associated with the number of symptoms reported (F(2, 158) = 6.3, p<.01) (see Figure). Neither the total score on the spiritual belief construct nor the support network construct was associated with the number of troublesome symptoms reported by patients.

Further analysis of the type of symptom reported by patients revealed additional interesting findings. Types of symptoms were categorized into the following three groups: no symptoms, pain-related symptoms (e.g., chest pain, stomach pain, pain in extremities), and other types of troublesome symptoms (e.g., lack of appetite, trouble sleeping, dyspnea). For the spiritual beliefs construct, 67.7% of responses associated with pain-related symptoms fell below the median score compared to only 46.6% and 49.3% of the "no symptoms" and "other symptoms" groups, respectively. The same pattern was identified for the support network construct: 64.5% of responses in the pain category fell below the median score compared to 47.8% and 48.7% of the "no symptoms" and "other symptoms" categories. These overall results point to the potential association between symptom reduction - and particularly, pain management - and QOL, and represent a point of intervention for clinicians.

It is interesting to note that none of the biomarkers (i.e., hemoglobin, albumin, and Kt/V) was significantly associated with the patient's QOL, including both the spiritual beliefs and support network constructs. In fact, the only significant association between physical functioning and QOL was found between the Karnofsky rating and overall score on the SWLS (r(160)=.17, p<.05).

The results of this study are promising and indicate potential points of intervention for improving the QOL among dialysis patients. Better assessment and treatment of patients' symptoms would seem to have a very positive effect on ESRD patients' QOL. Since pain was the most frequently reported symptom, better pain management for ESRD patients appears to be a priority to improve ESRD patient QOL. In addition, the lack of any significant association between the various biomarkers studied and quality of life underscores the relative importance of psychosocial and spiritual dimensions from the patient's perspective. The results also indicate the utility of a brief, 9-item scale that can be used to document the patient's spiritual beliefs and support network. Although more research is needed to further validate this scale, when administered in conjunction with other QOL scales, it can provide a more comprehensive assessment of a patient's psychosocial needs.

End Notes

[i] Cohen R, Mount BM, Strobel MG, Bui F. The McGill Quality of Life Questionnaire: A Measure of Quality of Life Appropriate for People with Advanced Disease. Pall Med 1995; 9: 207-219.

[ii] Diener E, Emmons R, Larsen J, Griffin, S. The Satisfaction with Life Scale. J Personality Assessment 1985; 49 (1): 71-75.

[iii] RAND, KDQOL-SFTM User's Manual, 1993.

[iv] DiMatteo MR, Hays R. The Significance of Patients' Perceptions of Physician Conduct: A Study of Patient Satisfaction in a Family Practice Center. J Community Health 1980; 6: 18-34.

[v] Karnofsky DA, Burchenal JH. The Clinical Evaluation of Chemotherapeutic Agents in Cancer, in Evaluation of Chemotherapeutic Agents in Cancer, ed. Macleod CM. New York: Columbia University Press, 1949, pp 191-205.

<<< Previous Next >>> [ Go Up ]

Promoting Excellence in End-of-Life Care was a national program of the Robert Wood Johnson Foundation dedicated to long-term changes in health care institutions to substantially improve care for dying people and their families. Visit for more resources.

Promoting Excellence logo and link to home page