KDQOL Frequently Asked Questions

Administering the KDQOL

Question: Is it valid to use the Symptoms/Problems scale and the Effects of Kidney disease scale only?

Response: You can use the symptoms/problems and the effects of kidney disease subscales of the KDQOL a la carte if you prefer. As you note, this means you will be only measuring part of what is captured by the KDQOL survey. We suggest you read the following article: Rao, S., Carter, W.B., Mapes, D.L., Kallich, J.D., Kamberg, C.J., Spritzer, K.L., & Hays, R.D. (2000). Development of subscales from the symptom/problems and effects of kidney-disease items in the Kidney Disease Quality of Life (KDQOL™) Instrument. Clinical Therapeutics, 22, 1099-1111.

Question: I moved the question on sexuality so it was not the last question because it is uncomfortable to end on that note. I think the question flows better about the symptoms questions. Can one change the order in which the questions of the KDQOL are administered?

Response: The KDQOL-36 item on sex life appears at question 35 (just prior to the last question; see item below). The KDQOL 1.3 questions on sexual activity appear as question 16, 16a and 16b.

Whenever you change a standardized measure, including changing the order of questions, there is some risk of problems related to comparability with the original measure. The effects are likely to depend on the specific situation but order effects, when statistically significant, are generally small in magnitude. See, e.g., http://www.jstor.org/stable/4038782.

In short, you are at risk when you compare your results to those from standardized survey administration because it is possible that the changes you implemented are responsible for any differences you see. If you feel the benefits of changing the order of items outweighs the potential risks, then your decision to alter the standardized survey may be the right one for your situation.

Question: I need to use the Kidney Disease Quality Of Life (KDQOL) questionnaire, but the following questions are not appropriate.
Q-16 due to cultural barriers,
Q-32 because it is not applicable question in our country,
Q-35 as health insurance is not a prevalent practice in our country,
Q-36 is not according to our currency.

Response: There is not a single KDQOL questionnaire. There are multiple surveys with differing number of items. We suspect from your reference to item numbers that you are looking at the KDQOL-SF version 1.3. Q-32, Q-35 and Q-36 are part of the optional demographic items; hence, you do not need permission to omit them.

Question: We are unsure what do about items not answered by patients completing the KDQOL.

Response: As indicated on page 5 of the KDQOL-SF 1.3 scoring manual, “Items that are left blank (missing data) are not taken into account when calculating the scale scores. Hence, scale scores represent the average for all items in the scale that the respondent answered.”

Question: The KDQOL-SF™ v 1.3 scoring manual includes demographic items. Do we have to administer them?

Response: The demographic items are optional and were created for administration in the U.S. These items may not be appropriate for your use of the KDQOL survey instrument.

Question: Do we need permission to use the KDQOL instrument? Is there a charge for its use?

Response: All of the surveys and tools are public documents, available without charge (for non-commercial purposes).

Please provide an appropriate citation when using these products. In some cases, the materials themselves include specific instructions for citation. For more information, see http://www.rand.org/health/surveys_tools/about_permissions.html

Question: Is it a violation of copyright to place a sticker on the front page of the KDQOL survey (for example: to document the patient’s case ID number)?

Response: No - it is not a violation of copyright.

Question: Can the KDQOL™ be used in patients who have chronic kidney disease (CKD), but have not yet started dialysis?

Response: Multiple users have used the KDQOL-1.3™ for pre-dialysis patients by excluding the questions about problems with access site (item 14L for hemodialysis) and catheter site (item 14M for peritioneal). In addition, the questions about dialysis staff encouragement and support (items 24A and 24B) are not applicable and should be omitted. Some have decided to administer the question about satisfaction with care (item 23) by changing “kidney dialysis” to “kidney disease.”

If using the KDQOL-36, you would also delete the questions about problems with access site (item 28a for hemodialysis) and catheter site (item 28b for peritoneal).

Users may want to do the same thing in using the KDQOL with patients who have received a kidney transplant.

Question: When administering the KDQOL, social workers understand not to ask leading questions, but find the test time consuming due the number of patients who cannot self-administer (vision or dependency) and the need to read the choices repetitively. Patients also appear to agonize over how to answer questions especially since questions ask for status over a month timeframe.

Is it OK for the person administering the test to advise patients to complete each question with the first answer that comes to mind?

Response: Yes, in fact we recommend that this be done.

Question: What is your recommendation about frequency of administering the KDQOL?

Response: This depends on the reason for the administration and the status of the patient. If it is for routine monitoring of a patient who is relatively stable, then 2-4 times a year (quarterly or every 6 months) is probably reasonable. If the patient is new to dialysis or has had changes in medication or dialysis prescription, then more frequent assessment might be warranted.

Question: Do you know whether there has been experience administering the KDQOL™ of the SF-36™ by telephone. I am researching the possibility of administering the SF-36™ to ESRD patients by phone.

Response: The SF-36™ has been administered by telephone in several studies including a study of the general US population (McHorney, Kosinski, & Ware, 1994, Medical Care, 32: 551-567). This study revealed that telephone data collection yielded lower rates of missing data, but more positive health-related quality of life scores (suggesting some social desirable response bias by telephone) Data collection costs for the telephone were 77% higher than for the mail survey.

Scoring & Interpreting the KDQOL

Question: I used the KDQOL in my thesis and I have a question about the scoring: can I average all the sub-scales of the questionnaire to get a one total score?

Response: The “KDQOL” is not a single instrument. There is the original long form that included 134 survey items, the subsequent short form (KDQOL-SF 1.3), and then an even shorter short form (KDQOL-36). The latter yields scores for the SF-12 physical and mental health summary scores plus the burden of kidney disease, symptoms/problems of kidney disease, and effects of kidney disease. You could create a total score but it is unclear what it would represent and we don’t recommend it. However, some investigators have created a Kidney Disease Component Summary (KDCS) that you may want to consider using: http://www.ncbi.nlm.nih.gov/pubmed/20973987 .

Question: Can we use SPSS instead of EXCEL to score the KDQOL?

Response: You can use whatever program you prefer to score the KDQOL. We provide EXCEL spreadsheets for scoring on the KDQOL website. There is also a SAS option.

Question: The Measures Assessment Tool used by ESRD surveyors when they survey dialysis clinics expects facilities to report KDQOL-36 scores and to assign those scores a risk level. I’ve heard that I should compare the patient’s score to the DOPPS age x gender chart (see “Downloads” page for chart). How can I tell risk level (low, average, high) from the DOPPS data?

Response: According to staff at Arbor Research Collaborative for Health that conducted the Dialysis Outcomes and Practice Patterns (DOPPS) research, scores within one standard deviation indicated “average” risk when adjusted by age and gender. Scores below one standard deviation from the mean indicate higher risk. Scores above than one standard deviation from the mean indicate lower than average risk.

More information on the DOPPS study can be found in:
Mapes DL, Bragg-Gresham JL, Bommer J, Fukuhara S, McKevitt P, Wikstrom B, & Lopes AA. (2004). Health-related quality of life in the Dialysis Outcomes and Practice Study (DOPPS). Am J Kidney Dis, 2004 Nov; 44(5 Suppl 2):54-60.

Question: What does it mean when the mean less the standard deviation on the DOPPS chart is a negative number? This happened with some of my female patients.

Response: According to staff at Arbor Research Collaborative for Health that conducted the Dialysis Outcomes and Practice Patterns (DOPPS) research, the mean for some age groups and genders of dialysis patients was quite low. Any score within one standard deviation of the mean indicates average risk when adjusted by age and gender. More information on the DOPPS study can be found in the citation above.

Question: I incorrectly entered a response for a question the patient did not answer and now I can’t get the KDQOL-36 Excel scoring template to give me the patient’s scores. What am I doing wrong?

Response: The spreadsheet requires missing data to be represented as a period “.”

Question: How should one score the KDQOL™ when a subject checks off two answers for a single question?

Response: It might make sense to average the responses or to randomly pick one over the other. It definitely doesn’t make sense to systematically pick either one favorable or less favorable. If the responses are not adjacent to one another in a large way (say 1 and 5, for example) then you might consider whether these are better coded as missing data (indicative of such poor quality response that neither is believable).

Because of the way the scoring is done for the SF portion of the instrument, it would be better to randomly pick one or the other of two responses, although you could use the average if the average comes out to a whole number. Again, you may wish to consider coding them as missing if the responses are not adjacent to each other in a large way.

Question: For patients with CKD not on dialysis, if we exclude the questions about problems with access site (item 14L for hemodialysis) and catheter site (item 14M for peritioneal), do we need to adjust the scoring algorithm or should we just consider this item to be missing for these patients?

Response: You should consider this item missing and be cautious about comparing symptom/problems scale scores from your sample to results for persons on dialysis.

Question: Are there national norms for the KDQOL-SF™ 1.3?

Response: There are national norms for the SF-36 and RAND-36 (see e.g.; Hays, Prince-Embury, & Chen, 1998). Norms for the kidney disease-targeted scales gathered in Arbor Research Collaborative for Health's Dialysis and Outcomes Practice Patterns Study (DOPPS) are available from http://www.dopps.org/DPM/.

Question: Are there other options for scoring and analysis in addition to HDO,Inc.? People who have gotten their information are concerned about cost.

Response: Users can score and analyze the KDQOL themselves following the user manual instructions. We have also provided on this site an Excel program for scoring the KDQOL. Alternatively, they might establish a relationship with a local college or university where there might be interest among students or faculty in using the data for research or educational purposes.


Question: Where can I find more information about the translation and evaluation of the KDQOL surveys?

Response: All of the information regarding the translated KDQOL surveys is located on our main KDQOL page. The translations have been created wholly by the listed contributors, and have not been evaluated or reviewed by RAND. RAND is providing the translations as a courtesy, and is not responsible for any inaccuracies or errors in the documents below. RAND does not have any translation certifications available. If you are interested in translating any surveys into another language, see our translation guidelines.

Question: How does the KDQOL-36™ compare with the KDQOL1.3™?

Response: The KDQOL-36™ is a subset of the KDQOL1.3™, so the 1.3 manual still applies. A crosswalk of the item numbers are as follows:

1 1 (items 1-12 correspond to SF-12)
3b 2
3d 3
4b 4
4c 5
5b 6
5c 7
8 8
9d 9
9e 10
9f 11
10 12
12a-12d 13-16 (burden of kidney disease)
14a-14l/m 17-28a/b (symptoms of kidney disease)
15a-15h 29-36 (effects of kidney disease)

Question: How does Version 1.3 of the KDQOL compare with Version 1.2?
Response: KDQOL™ Version 1.3 “differs from KDQOL-SF™ 1.2 by adding a screening item about sexual activity” (Hays et al., 1997, P-7994, p. 2).

Question: How do you answer the statement from those that developed the SF-36 that the RAND-36 led to higher scores than the SF-36?

Response: We point out the differences in scoring and the rationale for the RAND-36 scoring (see answer to next question).

Question: What is the difference between the SF-36™ and the RAND-36™?

Response: The SF-36™ and RAND-36™ include the same set of items that were developed in the Medical Outcomes Study. Scoring of the general health and pain scales is different, however. The NEMC scoring is used to provide a direct comparison with general population data that NEMC collected. However, the NEMC scoring system for one of the general health items and the pain scale is inconsistent with the simple summated scoring used for the other SF-36 items.

Specifically, the Excellent to Poor general health item is scored differently by NEMC. On the 0-100 scoring system, NEMC assigns Very Good the value of 85 instead of 75, and Good the value of 60 instead of 50. One of the pain items is scored conditional on the value of another by NEMC, and this increases the correlation between the two items. If one estimates internal consistency reliability for the pain scale using the NEMC scoring method, the coefficient will be artificially inflated. The differences in scoring are summarized by Hays, Sherbourne, and Mazel (Health Economics, 2: 217-227, 1993).

Question: When should the physical and mental health summary scales from the SF-36™ recommended by NEMC be used?

Response: We recommend that the SF-36™ physical and mental summary scores be used if a comparison to the NEMC reference values is desired. However, the scoring recommended by NEMC is based on a model that is inconsistent with observed data—it forces the correlation between physical and mental health scores to be zero, but the true correlation is positive and statistically significant (Hays, Marshall, et al., Journal of Consulting and Clinical Psychology, 62: 441-449, 1994). Therefore, we recommend that users also derive physical and mental health summary scores using an oblique (correlated) factor model (Farivar, Cunningham, & Hays. Health and Quality of Life Outcomes, 5: 54, 2007).

Question: Is there any reason why you wouldn’t keep KDQOL™ scores in the medical record? It seems that having the results here would make the scores more useful to all members of the team.

Response: We believe that having the scores in the medical record is an exciting possibility for maximizing the usefulness of health-related quality of life data in clinical practice.