CORRELATION BETWEEN 2 INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS

0022-5347/05/1733-0835/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 173, 835– 840, March 2005 Printed in U.S.A.

DOI: 10.1097/01.ju.0000152672.83393.61

Infection/Inflammation CORRELATION BETWEEN 2 INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS ERENEE SIRINIAN,* KATHRYN AZEVEDO

AND

CHRISTOPHER K. PAYNE†

From the Stanford University Medical Center, Stanford, California

ABSTRACT

Purpose: Two statistically validated interstitial cystitis (IC) symptom instruments have been used in clinical research. In this study we report what is to our knowledge the first direct comparison of the University of Wisconsin Symptom Instrument (UWI) with the O’Leary-Sant instruments, that is the IC Symptom Index (ICSI) and IC Problem Index (ICPI). Materials and Methods: A convenience sampling of 107 patients with IC evaluated at the urology clinic at our institution from March through August 2000 were asked to complete ICSI, ICPI and UWI. The scores were analyzed and graphed. Results: Mean ICSI, ICPI and UWI scores ⫾ SD were 12.6 ⫾ 4.3, 10.5 ⫾ 3.8 and 25.5 ⫾ 10.8, respectively. UWI and ICSI scores correlated at r ⫽ 0.80 (p ⬍0.01) and ICSI/ICPI scores correlated at r ⫽ 0.83 (p ⬍0.01). ICSI and ICPI contain 4 symptom-problem pairs. There are relevant differences in the correlation of symptom severity (ICSI question) and its associated problem (ICPI question). The symptom of urgency correlated best with the associated problem (r ⫽ 0.84), followed by nocturia (r ⫽ 0.82), pain (r ⫽ 0.70) and frequency (r ⫽ 0.68). Conclusions: The ICSI and UWI symptom instruments correlate strongly in a large population of patients with IC. The data presented will aid in interpreting the literature. A relatively poor correlation between pain symptom-problem pairs in the O’Leary-Sant instruments is probably an artifact of wording. The word pressure appears in ICPI but not in ICSI. We suggest using parallel wording of the pain symptom-problem pair containing the identical phrase, burning, pain, discomfort or pressure, to improve the ICSI/ICPI correlation and more accurately reflect the clinical condition. KEY WORDS: bladder; cystitis, interstitial; questionnaires; signs and symptoms; pain

Interstitial cystitis (IC) is a painful bladder disorder of unknown etiology. Because to our knowledge there is currently no ideal diagnostic test, IC remains a diagnosis of exclusion. Assessment of a patient with IC symptoms includes what studies may be necessary to rule out other diseases plus a targeted evaluation of bladder function. Ultimately the diagnosis rests on the overall picture and largely on patient symptoms. Current treatment focuses on decreasing symptoms and achieving remission when possible, but there are few objective tools that are useful for following disease activity and response to treatment. In this type of disease the assessment of symptoms is a relevant and even critical clinical concern. Reliable symptom evaluation is difficult in any disease and especially so in IC. Symptoms vary widely among cases and fluctuate in severity with time. To study the efficacy of IC treatments a reproducible, dependable method of evaluating symptoms is required. Urinary frequency and bladder capacity can be evaluated using voiding diaries and cystometry. These tools do not capture the full range of IC symptoms. In particular, they do not adequately assess the severity and impact of pain. Pain is often the key symptom, which makes quantifying symptoms especially problematic.

Structured instruments are required to assess the severity of patient symptoms and the impact or problem associated with the symptom. Various instruments have been used in IC clinical practice and research. Likert scales have been used in several studies.1–3 The single most commonly used outcome measure in clinical trials has been the Patients Overall Rating of Improvement of Symptoms question that asks patients about their overall level of improvement.4 It has been argued that the similar but structurally symmetrical Global Response Assessment is a superior scale.5 In addition to symptoms, the impact of a disease on quality of life (QOL) is an important outcome measure. Generic QOL instruments such as the RAND 36-Item Health Survey may be used to assess how any chronic condition such as IC affects patient ability to fulfill life roles.6 However, although general scales are attractive because of broad applicability, they can lead to incorrect conclusions due to inadequate sensitivity.7 This is a critical issue with IC because of the many associated diseases. Disease specific QOL instruments are preferred because patient health status data are linked to familiar clinical information, facilitating clinical interpretation of the information.5 Unfortunately to our knowledge there are as yet no disease specific QOL instruments for IC. Two groups recognized the need to develop broad symptom indexes specifically for IC. Keller et al developed the University of Wisconsin Index (UWI),8 and O’Leary et al developed

Submitted for publication May 11, 2004. * Current address: Chicago College of Osteopathic Medicine, Midwestern University, Chicago, Illinois. † Financial interest and/or other relationship with Yamanouchi, American Medical Systems, ICOS and Schwarz. 835

836

CORRELATION OF INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS

FIG. 1. UWI and ICSI total scores correlated at r ⫽ 0.80, p ⬍0.01 (95% CI 0.69 to 1.7)

the O’Leary-Sant IC Symptom Index (ICSI) and IC Problem Index (ICPI).9 These statistically validated instruments use patient history to assess symptom and/or problem severity. Keller et al started from an extensive literature review and in depth interviews with patients and clinicians to determine key disease symptoms.8 Sample questions were then evaluated by a small group of patients. The final UWI includes a core of 7 IC related questions about frequency, urgency, nocturia and pain. These items are mixed in with 18 reference items about other medical problems, eg shortness of breath, back pain, headaches, etc. Although the original sample size of 17 subjects was small, Keller et al noted that the scale seemed to measure important IC symptoms. Patients with IC scored much higher than controls on IC specific items but not on reference items. Internal consistency was high. When the scale was formally evaluated in 1998, the most attractive feature of the scale was its clinically apparent face validity and ease of implementation.10 O’Leary et al used experience with more than 400 patients during a 10-year experience at an IC clinic at New England Medical Center and a robust statistical validation process to develop ICSI and ICPI.9 It was believed that the tool might be useful for the evaluation and management of IC cases, and it would facilitate clinical research.11 Each scale contains 1 question about each of 4 symptoms, namely nocturia, frequency, urgency and bladder pain. The 2 scales separate the evaluation of symptom severity from that of the impact or problem caused by the symptoms. ICSI has also been demonstrated to be responsive to changes in patient condition.12 There is limited experience with these IC specific symptom instruments in clinical research. Researchers and clinicians have typically used 1 or another of these indexes but to our knowledge the 2 have not been directly compared. For example, ICSI and ICPI were used to evaluate oral treatment with suplaplast to sulate (IPD-1151T), Elmiron (Ortho-McNeil Pharmaceutical, Raritan, New Jersey), resiniferatoxin and

bacillus Calmette-Guerin.12, 13 UWI has been used to evaluate improvement in IC symptoms with bacillus CalmetteGuerin treatments.14 In this analysis we sought further to understand if patient reporting of IC symptoms with these 2 validated scales is comparable. MATERIALS AND METHODS

Inclusion criteria. A convenience sampling of 107 patients with IC presenting from March 2000 through August 2000 at the department of urology at our institution were asked to complete ICSI, ICPI and UWI while waiting to be seen by the urologist. This population encompassed a broad spectrum of disease, including severe and mild disease in newly diagnosed and chronic cases. These cases would be best described by the current International Continence Society definition of painful bladder syndrome as “the complaint of suprapubic pain related to bladder filling, accompanied by other symptoms such as increased daytime and night-time frequency, in the absence of proven urinary infection or other obvious pathology,”15 because the level of evaluation was widely variable. Some patients had undergone a detailed evaluation, including cystoscopy under anesthesia, urodynamics and potassium sensitivity testing, while others had been evaluated by history, physical examination and urinalysis without invasive testing. We excluded patients from this study if there were any diseases listed in National Institute of Diabetes and Digestive and Kidney Disease exclusion criteria or any other significant factors in the history or symptom complex that raised the possibility of another primary cause for patient symptoms. This patient population represents a realistic, broad spectrum of those seeking treatment for IC. These relaxed entry criteria were representative of how symptom instruments are used in clinical practice. UWI symptom instrument. The UWI includes 7 IC symptom items, namely bladder discomfort, bladder pain, getting

CORRELATION OF INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS

837

FIG. 2. ICSI and ICPI total scores correlated at r ⫽ 0.83, p ⬍0.01 (95% CI 0.65 to 0.84)

up at night to go to the bathroom, going to the bathroom frequently during the day, difficulty sleeping because of bladder symptoms, urination urgency and a burning sensation in bladder. Each is scored between 0 —not at all to 6 —a lot (Appendix 1). The patient is instructed to circle the “one number answer that comes closest to the way” the patient feels that day. The total score is 0 to 42. ICSI and ICPI instruments. This questionnaire is actually 2 instruments (Appendix 2), each containing 4 items, namely frequency, urgency, nocturia and bladder pain. The questions ask about the overall level of severity for each symptom (ICSI) and the amount of problem caused by each symptom (ICPI) during the preceding month. The symptoms are scored from 0 to 5 with a maximum score of 20 and problems are scored from 0 to 4 with a maximum score of 16. Statistics. Using the statistical program SPSS 12.0 (SPSS, Chicago, Illinois) the scores were correlated to obtain the Pearson correlation coefficient, analyzed and then graphed.

totals. Mean ICSI, ICPI and UWI total scores in our population were 12.6 ⫾ 4.3, 10.5 ⫾ 3.8 and 25.5 ⫾ 10.8, respectively. The 2 symptom indexes, ICSI and UWI, correlated well at r ⫽ 0.80 (p ⬍0.01). Figure 1 shows this comparison graphically. One might expect that the problem index (ICPI) would correlate less well with the symptom indexes but results were similarly well correlated (figs. 2 and 3). ICSI and ICPI correlation among individual values. To understand better the relationship between individual IC symptoms and associated problems we then analyzed each of the 4 symptom-problem pairs in ICSI/ICPI. Table 1 lists mean scores and SDs for each symptom-problem pair. The symptom-problem pairs of urgency and nocturia correlated more strongly than the symptom-problem pairs of pain and frequent urination. Correlations between the UWI questions and the relevant items on the ICSI and ICPI were less strong (table 2). Three different UWI items refer in some way to pain.

RESULTS

DISCUSSION

Population demographics. Nine men (8%) and 98 women (92%) participated in the study. The average age of all patients was 46.0 years (range 21 to 74). The average age of women was 45.5 years (range 21 to 73) and the average age of men was 51.5 years (range 27 to 74). Of the patients 23 (22%) had a race or ethnicity other than white American, that is Asian in 9, Hispanic in 8, Indian in 2, and black American, Russian, French and Polish in 1 each. Means and correlation values among ICSI, ICPI and UWI

There is a strong correlation between the O’Leary-Sant symptom index and the UWI (r ⫽ 0.80), stronger than one might expect, given that 1 instrument asks about symptoms during the last month and 1 asks about the last day. Figure 1 provides a means to compare the total symptom score from among instruments and among populations. The graph allows clinicians who may primarily use 1 instrument to understand better the results of a study that reports outcome with the other instrument.

838

CORRELATION OF INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS

FIG. 3. UWI and ICPI total scores correlated at r ⫽ 0.79, p ⬍0.01 (95% CI 0.57 to 1.7)

We found that the O’Leary-Sant symptom and problem totals correlated highly (r ⫽ 0.83). However, the 4 individual symptom-problem pairs did not seem to perform equally well. Urgency and nocturia correlated much more strongly than pain and frequency. We hypothesize that 2 explanations are responsible for the relatively lower correlation among ICSIICPI item pairs measuring frequency and pain than those measuring urgency and nocturia. We believe that the lower correlation detected in the severity of urinary frequency and the problem caused by frequency is valid. Patients perceive the problem of urinary frequency differently depending on individual life-style factors. For example, a patient working indoors with easy access to a bathroom is less troubled by frequent urination than one working outdoors with few break periods. Therefore, the same degree of frequency could represent a much greater problem to some patients than to others, producing the observed variability. In contrast, we suspect that the difference between patient symptom and problem scores for pain is primarily due to an artifact caused by the omission of the words pressure and

TABLE 1. ICSI and ICPI individual paired question correlations Symptom/Problem Question ICSI-1 ICSI-2 ICSI-3 ICSI-4

to to to to

ICPI-3 ICPI-1 ICPI-2 ICPI-4

(urgency) (frequency) (nocturia) (pain)

Correlation 0.84 0.68 0.82 0.70

discomfort from the symptom question. Pain appears in the problem question as “burning, discomfort, pain, or pressure.” We noted that many patients with IC respond to the word pressure in ICPI, while denying pain or burning. Patient comments written in the ICSI and ICPI margins next to their scores indicated confusion in the definition of pain. In at least 1 clinical study that used the O’Leary-Sant ICSI-ICPI as a model for the symptom instrument the wording on symptoms severity was changed to address the wide variation in how patients with IC perceive pain, that is, “How often have you had pain (burning, aching) or discomfort (pressure, cramps) in your lower abdomen, bladder, urethra or vagina?”2 TABLE 2. Correlations of UWI question with relevant ICSI and ICPI items UWI Question (No.)

Mean Total Score ⫾ SD

Respective Question Correlation ICSI

ICPI

Bladder discomfort (1) 4.3 ⫾ 1.6 0.55 0.55 Bladder pain (2) 3.6 ⫾ 2.0 0.75 0.73 Nocturia (10) 3.7 ⫾ 2.1 0.85 0.82 Frequency during day (18) 4.2 ⫾ 1.7 0.61 0.73 Sleeping problem from nocturia (21) 3.3 ⫾ 2.3 0.76 0.78 Urgency (23) 3.9 ⫾ 1.9 0.48 0.52 Burning (25) 3.0 ⫾ 2.3 0.66 0.61 The correlation between each UWI symptom score (questions 1, 2, 10, 18, 23 and 25) with the respective ICSI score was 0.79 and the correlation between the 1 UWI problem (question 21) score and the total ICPI score was 0.67.

839

CORRELATION OF INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS

Because these instruments are used in clinical and research applications, our results are relevant to each one. Clinicians will be able easily to compare outcomes when reading the medical literature. Information about new treatments reported with ICSI or UWI can be applied to clinician practice. Similarly clinical trials may be interpreted more uniformly, so that the personal preferences of measuring instrument from 1 research group will not exclude those who want to apply their results to a different patient population with a different symptom instrument. Our study is limited by investigating only 1 point in time. We did not study how these instruments compare with changes in patient condition after treatment or with time. Our patient sample is likely much broader than that with which the instruments were developed, although we believe this to be a strength of the study in representing a crosssection of the patient population. Finally, since this study was initiated, a new IC symptom instrument has been introduced, namely the pelvic pain and urgency/frequency scale.16 This instrument emphasizes the symptom of pain by increasing the number of questions about pain. It was not developed or validated with traditional statis-

tical methods but it has been validated against a commonly used diagnostic tool, that is the potassium sensitivity test. To our knowledge the additional questions about pain have not yet been demonstrated to capture different and valuable information. It would be interesting to see how this scale performs in comparison to the 2 more established instruments. CONCLUSIONS

There is a strong correlation between UWI and ICSI. Clinicians and future researchers using 1 or another of these instruments can now compare scores using the graph provided when interpreting reports of new treatments. A high correlation exists between ICSI and ICPI total scores, and for the symptomproblem pairs of urgency and nocturia. There was a significant disparity between frequency symptoms and their associated problem. Life-style factors might explain the different impact of urinary frequency on an individual life. The relatively lower correlation between pain severity and the problem that it causes is likely due to an artifact of wording, which could probably be improved by including the word pressure in ICSI. We advocate this change for future clinical research.

APPENDIX 1: KELLER UNIVERSITY OF WISCONSIN SYMPTOM INSTRUMENT

Please circle the one number answer that comes closest to the way you feel, whether or not you have the following symptoms. Symptom

Not At All

1. Bladder Discomfort 2. Bladder Pain 3. Other Pelvic Discomfort 4. Headache 5. Backache 6. Dizziness 7. Feelings of Suffocation 8. Chest Pain 9. Ringing in Ears 10. Getting Up at Night to Go to the Bathroom 11. Aches in Joints 12. Swollen Ankles 13. Nasal Congestion 14. Flu 15. Abdominal Cramps 16. Numbness or Tingling in Fingers or Toes 17. Nausea 18. Going to the Bathroom frequently during the day 19. Blind Spots or Blurred Vision 20. Heart Pounding 21. Difficulty Sleeping because of Bladder Symptoms 22. Sore Throat 23. Urgency to Urinate 24. Coughing 25. Burning Sensation in Bladder

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

(Circle one number on each line) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

A Lot 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

APPENDIX 2: O’LEARY-SANT SYMPTOM AND PROBLEM INSTRUMENTS

Identifying Interstitial Cystitis To help your physician determine if you have Interstitial Cystitis, please put a check mark next to the most appropriate response to each of the questions below. Then add up the numbers to the left of the check marks and write the total below. Interstitial Cystitis Symptoms Index During the past month: How 0. 1. 2. 3. 4. 5.

often have you felt the strong need to urinate with little or no warning: Not at all Less than 1 time in 5 Less than half the time About half the time More than half the time Almost always

Interstitial Cystitis Problem Index During the past month how much has each of the following been a problem for you. Frequent urination during the day? 0. No problem 1. Very small problem 2. Small problem 3. Medium problem 4. Big problem Getting up at night to urinate?

Have you had to urinate less than 2 hours after you finished urinating? 0. Not at all 1. Less than 1 time in 5 2. Less than half the time 3. About half the time 4. More than half the time 5. Almost always

0. 1. 2. 3. 4.

No problem Very small problem Small problem Medium problem Big problem Need to urinate with little warning?

840

CORRELATION OF INTERSTITIAL CYSTITIS SYMPTOM INSTRUMENTS APPENDIX 2: CONTINUED

Interstitial Cystitis Symptoms Index During the past month: How 0. 1. 2. 3. 4. 5.

often did you most typically get up at night to urinate? Not at all Once per night 2 times per night 3 times per night 4 times per night 5 or more times per night

Interstitial Cystitis Problem Index During the past month how much has each of the following been a problem for you. Need to urinate with little warning? 0. No problem 1. Very small problem 2. Small problem 3. Medium problem 4. Big problem

Have you experienced pain or burning in your bladder? 0. Not at all 1. A few times 2. Fairly often 3. Usually 4. Almost always

Burning, pain, discomfort, or pressure in your bladder? 0. No problem 1. Very small problem 2. Small problem 3. Medium problem 4. Big problem

Add the numerical values of the checked entries: Total score

Add the numerical values of the checked entries: Total score

REFERENCES

1. Korting, G.-E., Smith, S. D., Wheeler, M. A., Weiss, R. M. and Foster, H. E., Jr.: A randomized double-blind trial of oral L-arginine treatment of interstitial cystitis. J Urol, 161: 558, 1999 2. Propert, K. J., Schaeffer, A. J., Brensinger, C. M., Kusek, J. W., Nyberg, L. M., Landis, J. R. et al: A prospective study of interstitial cystitis: results of longitudinal followup of the Interstitial Cystitis Data Base cohort. J Urol, 163: 1434, 2000 3. Mazurick, C. A., Landis, J. R. and the Interstitial Cystitis Data Base Study Group: Evaluation of repeat daily voiding measures in the National Interstitial Cystitis Data Base study. J Urol, 163: 1208, 2000 4. Parsons, L. C., Benson, G., Childs, S. J., Hanno, P., Sant, G. R. and Webster, G.: A quantitatively controlled method to study prospectively interstitial cystitis and demonstrate the efficacy of pentosanpolysulfate. J Urol, 150: 845, 1993 5. Propert, K. J., Payne, C., Kusek, J. W. and Nyberg, L. M.: Pitfalls in the design of clinical trials for interstitial cystitis. Urology, 60: 742, 2002 6. Wagner, A. K., Shea, C. C. and Sant, G. R.: Health-related quality-of-life issues in interstitial cystitis. In: Interstitial Cystitis. Edited by G. R. Sant. Philadelphia: Lippincott-Raven Publishers, chapt. V, p. 261, 1997 7. McDowell, I. and Newell, C.: Measuring Health. New York: Oxford University Press, 1987 8. Keller, M. L., McCarthy, D. O. and Neider, R. S.: Measurement of symptoms of interstitial cystitis. A pilot study. Urol Clin North Am, 21: 67, 1994 9. O’Leary, M. P., Sant, G. R., Fowler, F. J., Jr., Whitmore, K. E. and Spolarich-Kroll, J.: The interstitial cystitis symptom index

and problem index. Urology, suppl., 49: 58, 1997 10. Goin, J. E., Olaleye, D., Peters, K. M., Steinert, B., Habicht, K. and Wynant, G.: Psychometric analysis of the University of Wisconsin Interstitial Cystitis Scale: implications for use in randomized clinical trials. J Urol, 159: 1085, 1998 11. O’Leary, M. P. and Sant, G. R.: The interstitial cystitis symptom and problem indices: rationale, development, and application. In: Interstitial Cystitis. Edited by G. R. Sant. Philadelphia: Lippincott-Raven Publishers, chapt. V, p. 271, 1997 12. Lubeck, D. P., Whitmore, K., Sant, G. R., Alvarez-Horine, S. and Lai, C.: Psychometric validation of the O’Leary-Sant interstitial cystitis symptom index in a clinical trial of pentosan polysulfate sodium. Urology, suppl., 57: 62, 2001 13. Ueda, T., Takaki, M., Ogawa, O., Yamauchi, T. and Yoshimuro, N.: Improvement of interstitial cystitis symptoms and problems that developed during treatment with oral IPD-1151T. J Urol, 164: 1917, 2000 14. Peters, K. M., Diokno, A. C., Steinert, B. W. and Gonzalez, J. A.: The efficacy of intravesical bacillus Calmette-Guerin in the treatment of interstitial cystitis: long-term followup. J Urol, 159: 1483, 1998 15. Abrams, P., Cardozo, L., Fall, M., Griffiths, D., Rosier, P., Ulmsten, U. et al: The standardisation of terminology of lower urinary tract function: report from the Standardisation Subcommittee of the International Continence Society. Neurourol Urodyn, 21: 167, 2002 16. Parsons, C. L., Dell, J., Stanford, E. J., Bullen, M., Kahn, B. S., Waxell, T. et al: Increased prevalence of interstitial cystitis: previously unrecognized urologic and gynecologic cases identified using a new symptom questionnaire and intravesical potassium sensitivity. Urology, 60: 573, 2002