Advertisement
American Journal of Kidney Diseases

Development of the Autosomal Dominant Polycystic Kidney Disease Impact Scale: A New Health-Related Quality-of-Life Instrument

Open AccessPublished:November 14, 2017DOI:https://doi.org/10.1053/j.ajkd.2017.08.020

      Background

      The impact of autosomal dominant polycystic kidney disease (ADPKD) on health-related quality of life (HRQoL) is not well understood due to a lack of instruments specific to the condition.

      Study Design

      Content for a new self-administered patient-reported outcome (PRO) questionnaire to assess ADPKD-related HRQoL was developed through clinical expert and patient focus group discussions. The new PRO instrument was administered to study patients with ADPKD to evaluate its reliability and validity.

      Setting & Participants

      1,674 adult patients with ADPKD participated in this research: 285 patients in focus groups to generate questionnaire content, 15 patients in debriefing interviews to refine the PRO questionnaire, and 1,374 patients to assess the performance and measurement properties of the PRO questionnaire.

      Outcome

      A new PRO questionnaire.

      Results

      The ADPKD Impact Scale (ADPKD-IS), consisting of 14 items representing 3 conceptual domains (physical, emotional, and fatigue) plus 4 additional questions, was developed. The instrument’s reliability (regarding internal consistency and test-retest consistency) and validity (content and construct) were supported.

      Limitations

      Need for more responsiveness testing when more data from clinical use become available over time. Complex concepts such as ADPKD-related pain and impact on a patient's HRQoL need further evaluation.

      Conclusions

      The ADPKD-IS is a new patient-centric tool that reliably and validly provides a standardized method for assessing HRQoL and overall disease burden in patients with ADPKD.

      Index Words

      Autosomal dominant polycystic kidney disease (ADPKD) is a rare hereditary systemic disease characterized by the progressive development of fluid-filled kidney cysts.
      • Torres V.E.
      • Harris P.C.
      • Pirson Y.
      Autosomal dominant polycystic kidney disease.
      • Grantham J.J.
      Autosomal dominant polycystic kidney disease.
      The expansion of the cysts eventually results in physically displacing and obstructing renal tubules, blood vessels, and lymphatics, as well as promoting apoptosis, atrophy, and fibrosis of the renal parenchyma.
      • Grantham J.J.
      • Mulamalla S.
      • Swenson-Fields K.I.
      Why kidneys fail in autosomal dominant polycystic kidney disease.
      ADPKD progresses at varying rates, resulting in loss of kidney function later in disease progression. Many patients with ADPKD present with symptoms such as hypertension, flank pain, urinary tract infection, nephrolithiasis, gross hematuria, or palpable kidneys. In later stages, the disease can have a broad spectrum of additional complications involving multiple organs.
      • Gabow P.A.
      Gabow PA 1996 PKD Chapter 13 Natural History ADPKD.pdf.
      • Torres V.E.
      • Harris P.C.
      Autosomal dominant polycystic kidney disease: the last 3 years.
      • Halvorson C.R.
      • Bremmer M.S.
      • Jacobs S.C.
      Polycystic kidney disease: inheritance, pathophysiology, prognosis, and treatment.
      An estimated 45% to 70% of patients with ADPKD progress to end-stage kidney disease by the age of 65 years,
      • Lentine K.L.
      • Xiao H.
      • Machnicki G.
      • Gheorghian A.
      • Schnitzler M.A.
      Renal function and healthcare costs in patients with polycystic kidney disease.
      at which point they may have cyst-filled kidneys weighing up to 30 pounds.

      National Kidney and Urologic Diseases Information Clearinghouse. Polycystic Kidney Disease. Bethesda, MD; 2007. https://www.niddk.nih.gov/-/media/FD324862F39546E88D22860766E3A9EA.ashx. Accessed September 23, 2016.

      ADPKD imposes significant burden on patients due to its complex symptomatology, hereditary nature, and gradual disease progression.
      • Tong A.
      • Rangan G.K.
      • Ruospo M.
      • et al.
      A painful inheritance–patient perspectives on living with polycystic kidney disease: thematic synthesis of qualitative research.
      ADPKD-related disease burden has been assessed in only a few studies and with instruments that have not been validated to assess the impacts of ADPKD.
      • de Barros B.P.
      • Nishiura J.L.
      • Heilberg I.P.
      • Kirsztajn G.M.
      Ansiedade, depressão e qualidade de vida em pacientes com glomerulonefrite familiar ou doença renal policística autossômica dominante.
      • Heiwe S.
      • Bjuke M.
      “An evil heritage”: interview study of pain and autosomal dominant polycystic kidney disease.
      • Miskulin D.C.
      • Abebe K.Z.
      • Chapman A.B.
      • et al.
      Health-related quality of life in patients with autosomal dominant polycystic kidney disease and CKD stages 1-4: a cross-sectional study.
      • Rizk D.
      • Jurkovitz C.
      • Veledar E.
      • et al.
      Quality of life in autosomal dominant polycystic kidney disease patients not yet on dialysis.
      General health assessment questionnaires, such as the 12-Item Short-Form Health Survey version 2 (SF-12v2),
      miss aspects of quality of life important to patients with ADPKD. As a result, these questionnaires fail to capture the totality of disease-related physical and mental impacts, making them less sensitive.
      • Miskulin D.C.
      • Abebe K.Z.
      • Chapman A.B.
      • et al.
      Health-related quality of life in patients with autosomal dominant polycystic kidney disease and CKD stages 1-4: a cross-sectional study.
      • Rizk D.
      • Jurkovitz C.
      • Veledar E.
      • et al.
      Quality of life in autosomal dominant polycystic kidney disease patients not yet on dialysis.
      Instruments specifically designed for kidney disease, such as the Kidney Disease Quality of Life (KDQOL) survey,
      • Hays R.D.
      • Kallich J.D.
      • Mapes D.L.
      • Coons S.J.
      • Carter W.B.
      Development of the Kidney Disease Quality of Life (KDQOLTM) instrument.
      do not capture burden before end-stage kidney disease in ADPKD, for which fatigue, pain, anxiety, nocturia, polyuria, and disruption of daily activities occur. Although the Brief Pain Inventory–Short Form (BPI-SF),
      • Cleeland C.S.
      • Ryan K.M.
      Pain assessment: global use of the Brief Pain Inventory.
      a measure of pain severity and pain interference with daily activities, has been used in kidney disease and end-stage kidney disease,
      • Davison S.N.
      Pain in hemodialysis patients: prevalence, cause, severity, and management.
      • Gamondi C.
      • Galli N.
      • Schönholzer C.
      • et al.
      Frequency and severity of pain and symptom distress among patients with chronic kidney disease receiving dialysis.
      it had not been evaluated for use in ADPKD. The ADPKD Impact Scale (ADPKD-IS) was developed to comprehensively assess health-related quality of life (HRQoL) in patients with ADPKD as a patient-centric valid measure.
      • Lecardeur L.
      • Joly D.
      [Quality of life in patients with autosomal dominant polycystic kidney disease] in French.

      Methods

      Overview

      In developing the ADPKD-IS instrument, we followed standard guidelines for development of a new patient-reported outcome (PRO) instrument.

      Food and Drug Administration. Guidance for Industry Use in Medical Product Development to Support Labeling Claims Guidance for Industry. 2009.

      • Patrick D.L.
      • Burke L.B.
      • Gwaltney C.J.
      • et al.
      Content validity—establishing and reporting the evidence in newly developed patient-reported outcomes (PRO) instruments for medical product evaluation: ISPOR PRO Good Research Practices Task Force Report: part 2—assessing respondent understanding.
      • Patrick D.L.
      • Burke L.B.
      • Gwaltney C.J.
      • et al.
      Content validity—establishing and reporting the evidence in newly developed patient-reported outcomes (pro) instruments for medical product evaluation: ISPOR PRO Good Research Practices Task Force Report: part 1—eliciting concepts for a new PRO instrument.
      Figure 1 provides an overview of the development process, which consisted of a series of individual studies. The New England Institutional Review Board (Needham, MA) served as the central review body for all studies included in this research.
      Figure thumbnail gr1
      Figure 1Development of the Autosomal Dominant Polycystic Kidney Disease Impact Scale (ADPKD-IS). Abbreviations: CKD, chronic kidney disease; PRO, patient-reported outcome.

      Recruitment

      Men and women 18 years or older with ADPKD were recruited through physician and family referrals, advocacy groups, and print advertising. Participating patients provided informed consent before any study-related activities.

      Development of a Conceptual Framework and Questionnaire

      A literature review that focused on disease-related unmet need and burden, PRO instruments, and key outcome gaps related to ADPKD was conducted. A list of categories (concepts) to be measured was compiled for further evaluation. Additional detail on the literature review is available in Item S1.
      Twenty-six ADPKD clinical experts from North America (n = 16), Europe (n = 8), and Japan (n = 2) were interviewed regarding: (1) the relevance of identified PRO instruments from the literature review, (2) issues likely considered burdensome to patients with ADPKD, (3) aspects of patients’ lives likely to improve following successful ADPKD treatment, and (4) potential concepts and questions (items) for an ADPKD-specific PRO instrument. The expert feedback was reviewed to identify issues that, from the perspective of the clinician, affected the HRQoL of patients with ADPKD.
      The burden of ADPKD on patients’ daily lives was explored in focus groups with 117 adult patients with ADPKD in chronic kidney disease (CKD) stages 1 to 5 from the United States, Turkey, Germany, United Kingdom, and Japan. Groups of up to 8 participants were single sex when possible due to urologic and body image topics. Patients discussed how ADPKD affected their general health, daily activities, physical or social activities, pain experience, urinary issues (urgency, frequency, and nocturia), and emotional well-being. Transcripts were coded, and concepts or themes mentioned by at least 2 participants in a group were considered relevant. Data saturation (the point at which additional sampling provides no new information) was achieved when no new concepts or themes were identified in subsequent groups.
      Based on information from patients and clinical experts, an ADPKD-specific conceptual framework (a model representing concepts/themes to be measured and their relationships) and an initial questionnaire were created. Fifteen cognitive debriefing interviews with US-based patients assessed the level of comprehension, understanding, and interpretation of all instructions and questioned wording and response options by the target audience (content validity), and the instrument was refined (question wording, format, and structure) via an iterative process. Additional content validation focus groups with 168 patients in Australia, Argentina, Brazil, Canada, China, Czech Republic, Hungary, Japan, Poland, Romania, South Korea, Spain, and the United States were conducted to further explore ADPKD-related pain and ensure global applicability of the measured concepts. Finally, the ADPKD-IS was reviewed by clinical experts, including physicians and nurses, to ensure that the instrument was a good reflection of the concepts to be measured (face validity).

      Quantitative Evaluation

      The ADPKD-IS was administered in an observational study (ClinicalTrial.gov study number NCT01430494) to obtain data for cross-sectional evaluation of its measurement properties (psychometrics) in 665 US-based patients with ADPKD and to evaluate clinical performance in 1,076 patients with ADPKD globally. The instrument’s stability at baseline and at 1 month was evaluated in a separate study with 298 US participants.

      Confirmatory Factor Analysis

      To evaluate construct validity (ie, how well a test measures what it claims to measure
      • Cronbach L.J.
      • Meehl P.E.
      Construct validity in psychological tests.
      ), data were subjected to confirmatory factor analysis to evaluate whether the hypothesized organization of questions in domains on the ADPKD-IS was empirically supported. Nonparametric and parametric tests were conducted to evaluate model fit (additional information on the specific tests can be found in Item S2). Iterative model refinements were performed using a randomly split data set (half used as a development sample and the other half used as a cross-validation sample).

      Item Response Theory

      After evaluating the assumption of local independence among items in each unidimensional APKD-IS construct, we estimated item response theory models to obtain item parameter estimates. Several 1- and 2-parameter models were compared to evaluate whether it was important to allow slopes to vary across items. Item category response curves and item/scale information functions were examined to evaluate the response categories for each item and detail the item and scale breadth of coverage for the latent state of ADPKD impact. Each domain was scored both as a simple sum-score and using item response theory parameters to apply differential weighting per item.

      Item-Level Psychometric Statistics

      Each item was examined to see whether it should be kept as part of its assigned domain in the conceptual framework. Within each domain, equality of item-total correlations was examined by testing the correlation difference between the mean of corrected item-total correlations (ie, correlation when the item is removed from the total score) and the most deviant corrected item correlation. Mean correlations for the scales were calculated using Fisher z transformation. Item-total correlations by item were calculated per domain, correcting for overlap of the actual item in the total score. Items with the smallest and largest variances on each scale were compared using Hartley Fmax test when any value > 3.0 suggests variability too discrepant for parametric approaches.
      • Keppel G.
      Design and Analysis: A Researcher’s Handbook.
      Sufficient item-total correlations were compared against a criterion of 0.40.
      • Martin B.
      • Pathak D.
      • Sharfman M.
      Validity and reliability of the Migraine-Specific Quality of Life Questionnaire (MSQ version 2.1).
      Each item was examined to ensure that its corrected item-total correlation for its own scale was significantly higher than that for either of the other scales. Correlation difference tests were used to compare the item-total correlations of each item with the other domains.

      Domain-Level Psychometric Statistics

      Cronbach internal consistency reliability (alpha) coefficients and their standard errors were estimated to evaluate the reliability of domain scores.
      • Cronbach L.
      Coefficient alpha and the internal structure of tests.
      • Nunnally J.C.
      • Bernstein I.H.
      Psychometric Theory.
      • Anastasi A.
      • Urbina S.
      Psychological Testing.
      • Duhachek A.
      • Iacobucci D.
      Alpha’s standard error (ASE): an accurate and precise confidence interval estimate.
      • Clark L.A.
      • Watson D.
      Constructing validity: basic issues in objective scale development.
      The average of all interitem correlations within a domain (item homogeneity) was also estimated.

      Convergent Correlations Between ADPKD-IS and Other Instruments

      Convergent validity (how the instrument relates to other instruments that are supposed to measure the same or similar concept) was assessed by correlating domain scores (both sum-score and item response theory–based score) with the SF-12v2 Physical Component Summary (PCS) and Mental Component Summary (MCS) scale scores and the BPI-SF
      • Cleeland C.S.
      • Ryan K.M.
      Pain assessment: global use of the Brief Pain Inventory.
      Pain Interference Score. The BPI-SF Pain Interference Score, a measure of how much that pain interferes with 7 daily activities, was chosen for evaluation because the ADPKD-IS assesses disease impact. Correlations > 0.40 were considered to be strong and indicative of the domains and questions measuring the same construct.
      • Cohen J.
      Statistical Power Analysis for the Behavioral Sciences.

      Test-Retest Reliability, Responsiveness, and Meaningful Difference

      The stability of domain scores was assessed using intraclass correlation coefficients over a 1-month test-retest interval in which correlations should exceed a threshold of 0.70.
      Correlations were examined between domain change scores and changes in SF-12v2 and BPI-SF scores.
      In the study sample, meaningful change was estimated using anchor-based methods by evaluating associations between ADPKD-IS domains and the concept measured by anchors, including a pain-based global change scale,
      • Juniper E.F.
      • Guyatt G.H.
      • Willan A.
      • Griffith L.E.
      Determining a minimal important change in a disease-specific Quality of Life Questionnaire.
      SF-12v2 PCS and MCS, and BPI-SF Pain Severity domain (a measure of worst, least, average, and current pain severity).

      Clinical Evaluation

      The clinical performance of the ADPKD-IS in patients with ADPKD with CKD stages 1 to 5 was assessed in 1,076 participants by evaluating sensitivity in the overall population and by CKD stage at baseline and in 6-month intervals up to 3 years.
      Mean scores and change from baseline were compared for the ADPKD-IS, SF-12v2, and EQ-5D (3-level version; EuroQol). Cohen d effect size was evaluated relative to patients with CKD stage 1.
      • Cohen J.
      Statistical Power Analysis for the Behavioral Sciences.

      Results

      Qualitative Development

      Literature review (4,801 articles, 27 relevant) identified several concepts relevant to HRQoL. The PRO instruments of potential interest due to ADPKD-relevant concepts (eg, kidney disease, pain, anxiety, depression, and urologic issues) or prior use in ADPKD studies were identified. However, there were no instruments specifically designed for or their validity evaluated in an ADPKD population. Identified concepts, including physical impact, emotional impact, and urinary concerns, and PRO instruments were discussed with 26 clinical experts from North America, Japan, and Europe. Clinical experts noted that patients with ADPKD typically are not affected in early disease stages, but with disease progression, concerns increase (eg, shortness of breath from pressure of the kidneys, difficulty eating, depression, and physical and emotional burden). Pain and discomfort were reported as a potential issue and recommended for probing with patients.
      Focus groups with 117 patients (United States = 42, Turkey = 15, Germany = 29, United Kingdom = 20, and Japan = 11) reflected the concepts generated from discussions with clinical experts and literature review. More than half the concepts were observed in the first focus group, and saturation was achieved by the fifth of the initial 20 concept-generating focus groups. Overall, 20 concepts were identified related to physical impact, emotional impact, urinary concerns, and other issues (Table 1). Examples of patient quotations are provided in Box 1 and Item S3. Contrary to clinical expert feedback, participants in all disease stages reported disease burden, but many accepted this as a normal state worsening with disease progression.
      Table 1ADPKD Focus Group Concepts
      ConceptFocus Groups Region
      United States (n = 42)Europe (n = 64)Japan (n = 11)
      Physical impact
       Impact on work/housework10 (23.8%)10 (15.6%)11 (100%)
       Limited functioning22 (52.4%)26 (40.6%)9 (81.8%)
       Self-care6 (14.3%)3 (4.7%)4 (36.4%)
       Effect on relationships (intimacy/sex)4 (9.5%)2 (3.1%)0 (0%)
       Pain/discomfort34 (81.0%)28 (43.8%)11 (100%)
       Pain affecting work/housework5 (11.9%)2 (3.1%)4 (36.4%)
       Pain with physical activity3 (7.1%)12 (18.8%)5 (45.5%)
       Lifestyle modification15 (35.7%)18 (28.1%)3 (27.3%)
      Emotional impact
       Fatigue36 (85.7%)29 (45.3%)3 (27.3%)
       Depression27 (64.3%)15 (23.4%)2 (18.2%)
       Anxiety31 (73.8%)30 (46.9%)2 (18.2%)
       Guilt15 (35.7%)12 (18.8%)6 (54.5%)
       Acceptance5 (11.9%)32 (50.0%)5 (45.5%)
      Urinary concerns
       Urinary urgency17 (40.5%)20 (31.3%)4 (36.4%)
       Urinary frequency17 (40.5%)27 (42.2%)3 (27.3%)
       Nocturia18 (42.9%)24 (37.5%)2 (18.2%)
      Other impacts
       Effect on diet5 (11.9%)32 (50.0%)7 (63.6%)
       Concern about body image22 (52.4%)25 (39.1%)10 (90.9%)
       Feeling thirsty10 (23.8%)19 (29.7%)4 (36.4%)
       Disruption of social/leisure activities26 (61.9%)24 (37.5%)11 (100%)
      Note: Values are given as number (percentage).
      Abbreviation: ADPKD, autosomal dominant polycystic kidney disease.
      Patient Quotations
      Impact on Daily Activities and Physical Limitations
      • “General housework, really, is quite difficult. I get tired and I have to keep sitting down. When my son was younger, I had trouble picking him up, because of the pain in the kidneys.”
      • “I have difficulties breathing sometimes, because my abdominal region is located fairly high up in my body and pushes up. It makes me pant sometimes.”
      • “I try to find shoes that don’t need to be tied up. I get ones where I can use a shoehorn only.”
      • “I used to do Aqua Fitness, too, but they give you these swim rings and I couldn’t wear them anymore, because of the pressure on my midriff.”
      • “It’s not an actual pain, but it’s a constant presence of something that’s not right. It’s led to my not doing things that I used to do every day, like cycling, which I loved doing.”
      Work and Financial Impact
      • “The pain only allows me to perform 50%. I always try to do my best, 100%. But when the pain comes, I can’t.”
      • “If it is a really bad pain day, you don’t want to go to work. You don't want to go out and do anything. You just want to take a bunch of Tylenol and stay in bed.”
      • “Even at work, walking across the office and back, I would always have shortness of breath and I always attributed that to PKD.”
      Emotional Impact
      • “It’s not just the intensity of pain. It can be relatively low-level, but it’s dragged out over a long period of time it wears you down and psychologically somebody can develop like a feeling of doom.”
      • “I think the psychological aspect is the most difficult one to bear and to come to terms with. The complications, the difficulties, the course and further course of the disease and all of that. The fear.”
      • “I know that there’s something in my body that’s not functioning or that may not one day function 100%. You'd like to think that you are invincible, so yeah, it’s a little bit scary. I think, oh well, my mom had an aneurysm. What if I’m going to have one too. It's just the unknown, I guess.”
      Abbreviation: PKD, polycystic kidney disease.
      Cognitive debriefing interviews with 15 US-based patients resulted in an instrument with 18 items grouped into 2 general domains: physical impact and social/emotional impact of ADPKD. In additional focus groups with 168 patients (North America = 31, South Korea = 16, Australia = 8, South America = 18, China = 15, Taiwan = 16, Japan = 12, Romania = 7, Spain = 16, Czech Republic = 15, Hungary = 6, and Poland = 8), initial concepts were confirmed as globally consistent. Patient focus groups showed general agreement in concepts between sexes and regions. Pain was described as complex, with 3 distinct types of pain (sharp acute, dull chronic, and fullness/discomfort) being relevant to patients with ADPKD.
      Nine model iterations were performed for confirmatory factor analysis. The hypothesized questionnaire organization was inconsistent with the empirically tested latent structure of the initial model. The best fit to the data was found using a bifactor model with 3 domains (Physical Debility, Fatigue, and Emotional Distress). The final model (Fig 2) showed no substantive differences between the parametric and nonparametric models, and fit statistics were similar between the developmental and cross-validation samples (Table 2). Analyses resulted in the removal of 4 items (coping, sleep, size/shape of abdomen, and frequent and/or urgent urination) from the domains. These domain-independent items were retained in the final questionnaire (Fig 2) due to strong endorsement by patients and relationship to some of the main disease features (eg, abdomen shape and urinary concerns).
      Figure thumbnail gr2
      Figure 2Autosomal Dominant Polycystic Kidney Disease Impact Scale (ADPKD-IS) final conceptual framework and model.
      Table 2ADPKD-IS Model Fit Statistics
      Sampleχ2dfCFINNFIRMSEARMSEA 90% CISRMR
      Parametric development split sample104.51590.9680.9500.0480.033-0.0630.032
      Parametric validation split sample109.67590.9700.9540.0510.036-0.0660.028
      Abbreviations: ADPKD-IS, Autosomal Dominant Polycystic Kidney Disease Impact Scale; CFI, comparative fit index; CI, confidence interval; NNFI, non-normed fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; χ2, Chi square.
      We also compared concepts measured by the SF-12v2 and ADPKD-IS (Fig 3) demonstrating a low level of overlap between the general instrument and the patient-centered measures of the ADPKD-IS.
      Figure thumbnail gr3
      Figure 3Concept mapping between the 12-Item Short Form Health Survey version 2 (SF-12v2) and Autosomal Dominant Polycystic Kidney Disease Impact Scale (ADPKD-IS).

      Measurement Properties

      Patients of all CKD stages were included in the cross-sectional study of 665 participants at baseline. Although earlier stages were more represented than later stages, a similar distribution of participants across disease stages was seen in the longitudinal study (Tables S1 and S2).

      Item Discrimination

      Using a 2-parameter logistic graded-response model, item discriminations (an item’s ability to distinguish among persons who have different levels of the trait being measured; ≥0.80 is recommended
      • Nunnally J.
      • Bernstein I.
      • Berge J.
      ) were found to range from 1.19 to 2.82 across scales (Table S3).
      The physical domain provided the greatest range of measurement of the target concept, while the fatigue domain provided the most restricted range of measurement. Compared to the SF-12v2, low levels of physical impact were less common in the study population, whereas low levels of emotional impact were more common.

      Item-Level Psychometric Statistics

      Correlations between individual item scores and overall domain scores ranged from 0.75 to 0.85 for physical, 0.86 to 0.90 for fatigue, and 0.51 to 0.81 for emotional, which are considered substantial and satisfactory to the hypothesized scale (Table S4). The lowest correlation was 0.70 for Item 13 (“feeling full before appetite was satisfied”). Hartley Fmax test (equality of variances for each item per scale) was 2.62 for physical, 1.24 for fatigue, and 1.53 for emotional.
      The percentage of patients reporting the lowest score (floor effects) on the 18 items ranged from 42% to 77%, whereas the percentage of patients reporting the highest score (ceiling effects) ranged from 1% to 6% per item (Table S5). Item-level skewness and kurtosis parameters likewise indicated a substantial departure from normality, with greater numbers of patients reporting lower scores on each item than higher scores.

      Domain-Level Psychometric Statistics

      Internal consistency of the domains was measured by coefficient alpha and average interitem correlation. Coefficient alphas exceeded the 0.70 reliability threshold for group comparisons for each domain: 0.94 for physical, 0.94 for fatigue, and 0.85 for emotional. The domains achieved appropriate average interitem correlations: 0.71 for physical, 0.84 for fatigue, and 0.58 for emotional. Minimal ceiling effects were observed per domain (0.5% physical, 2% fatigue, and 0.3% emotional). Observed floor effects were 44%, 41%, and 30% for physical, fatigue, and emotional, respectively (Table 3).
      Table 3ADPKD-IS Reliability
      Mean ± SDObserved Range% Floor% CeilingCoefficient AlphaAverage Interitem Correlation
      Overall score24.1 ± 11.714–7017.70.20.950.61
      Physical11.2 ± 6.07–3543.80.50.940.71
      Fatigue5.8 ± 3.43–1541.22.30.940.84
      Emotional7.2 ± 3.54–2030.10.30.850.58
      Abbreviations: ADPKD-IS, Autosomal Dominant Polycystic Kidney Disease Impact Scale; SD, standard deviation.

      Convergent Correlations With Other Instruments

      Correlations of domains with the SF-12v2 summary scores were all large: –0.68 for physical with the PCS, –0.58 for fatigue with the PCS, and –0.54 for emotional with the MCS, indicating that the domains measured related concepts (Table 4).
      Table 4ADPKD-IS Convergent Validity
      PhysicalFatigueEmotional
      Physical0.80920.7166
      Fatigue0.80920.6776
      Emotional0.71660.6776
      SF-12v2 PCS−0.6810−0.5757−0.4092
      SF-12v2 MCS−0.4798−0.5069−0.5362
      Note: Table shows product-moment correlations of ADPKD-IS domains with the SF-12v2 (convergent validity).
      Abbreviations: ADPKD-IS, Autosomal Dominant Polycystic Kidney Disease Impact Scale; MCS, Mental Component Summary; PCS, Physical Component Summary; SF-12v2, 12-Item Short Form Health Survey version 2.

      Test-Retest Reliability, Responsiveness, and Meaningful Difference

      Test-retest reliability coefficients (traditional correlations) for the 3 domains were high (0.89 for physical, 0.92 for fatigue, and 0.86 for emotional). Intraclass correlations between 3 and 4 weeks were consistent with this finding. Within-participant change scores observed for items and domain assessments indicated little change over the test-retest period, consistent with the slow gradual disease progression of ADPKD (Table 5).
      Table 5ADPKD-IS Responsiveness
      PhysicalFatigueEmotional
      Global rating of change in pain
      P0.0030.060.3
       Improved +1 (n = 15)−0.362 ± 0.81−0.444 ± 0.79−0.233 ± 0.50
       No change 0 (n = 83)−0.021 ± 0.43−0.120 ± 0.53−0.123 ± 0.62
       Worsened −1 (n = 10)0.443 ± 1.020.100 ± 0.740.125 ± 0.63
      BPI-SF Pain Severity
      P0.0010.230.3
       ≥2.0 points improved (n = 10)−0.128 ± 1.03−0.200 ± 1.22−0.275 ± 0.83
       −2 to +2 points; no change (n = 89)−0.082 ± 0.43−0.172 ± 0.49−0.124 ± 0.57
       ≤2 points worsened (n = 9)0.651 ± 0.930.185 ± 0.620.139 ± 0.68
      SF-12v2 PCS
      P0.90.20.2
       ≥3.0 points improved (n = 36)−0.036 ± 0.750.001 ± 0.680.021 ± 0.57
       −3 to +3 points; no change (n = 47)−0.042 ± 0.51−0.206 ± 0.56−0.133 ± 0.49
       ≤3 points worsened (n = 25)−0.023 ± 0.48−0.240 ± 0.56−0.280 ± 0.80
      SF-12v2 MCS
      P0.40.50.6
       ≥2.0 points improved (n = 28)0.066 ± 0.38−0.036 ± 0.68−0.045 ± 0.64
       −2 to +2 points; no change (n = 38)0.008 ± 0.41−0.193 ± 0.49−0.197 ± 0.58
       ≤2 points worsened (n = 42)−0.116 ± 0.81−0.175 ± 0.65−0.089 ± 0.60
      Note: Unless otherwise indicated, values are given as LS mean ± standard deviation.
      Abbreviations: ADPKD-IS, Autosomal Dominant Polycystic Kidney Disease Impact Scale; BPI-SF, Brief Pain Inventory Short Form; MCS, Mental Component Summary; PCS, Physical Component Summary; SF-12v2, 12-Item Short Form Health Survey version 2.
      Responder groups defined by pain showed significant differences on the ADPKD-IS physical domain (P = 0.003), with marginal differences on the Fatigue scale (P = 0.06). The impact of perceived change among 23% (n = 25) of the sample reporting any change was related to physical function, but not fatigue or emotional distress. Similarly, the BPI-SF Pain Severity scale demonstrated change related to physical function (P = 0.001), but not fatigue or emotional distress. More marked changes in pain severity on the BPI-SF appear to be associated with urinary symptoms on the ADPKD-IS. None of the domain assessments responded to change groups defined by criteria on the SF-12v2 PCS or SF-12v2 MCS (Table 5).

      Scoring

      For all items, each response option is assigned a value between 1 and 5. A score of 1 indicates “not difficult at all” or “not bothered at all.” A score of 5 indicates “extremely difficult” or “extremely bothered.” Each domain is scored by summing the 1 to 5 scores for the items in that domain (Fig 2) and dividing by the number of items completed.

      Clinical Evaluation

      Scores on the physical, emotional, and fatigue domains of the ADPKD-IS differed significantly between patients in CKD stage 3b versus CKD stage 1 at baseline. On the SF-12 PCS and EQ-5D index, differentiation from CKD stage 1 was observed as early as CKD stage 3a. The SF-12 MCS did not show differentiation. The SF-12 PCS and MCS scores were consistently above normal scores for a CKD population throughout all disease stages. SF-12 PCS score was above or at the level for the US general population through CKD stage 3a, whereas SF-12 MCS score was above or at the level for the US general population through CKD stage 4 (Fig 4). Percent change between CKD stages was more pronounced for the ADPKD-IS domains compared with other instruments (Fig 5). Using Cohen d effect size relative to CKD stage 1, the burden of disease could be distinguished as early as CKD stage 3 across all 3 domains of the ADPKD-IS, the SF-12vs PCS, and the EQ-5D, with the largest effect seen for CKD stage 5 across all scales (Fig 6).
      • Oberdhan D.
      • Krasa H.B.
      • Schaefer C.
      • Cole J.C.
      Health-related quality of life (HRQOL) measures in autosomal dominant polycystic kidney disease (ADPKD).
      Figure thumbnail gr4
      Figure 4Autosomal Dominant Polycystic Kidney Disease Impact Scale (ADPKD-IS), EQ-5D, and 12-Item Short Form Health Survey (SF-12) mean at baseline by chronic kidney disease (CKD) stage. Abbreviations: MCS, Mental Component Summary; PCS, Physical Component Summary.
      Figure thumbnail gr5
      Figure 5Percent difference compared to patients with chronic kidney disease stage 1 (CKD 1) at baseline. Abbreviations: ADPKD-IS, Autosomal Dominant Polycystic Kidney Disease Impact Scale; MCS, Mental Component Summary; PCS, Physical Component Summary; SF-12, 12-Item Short Form Health Survey.
      Figure thumbnail gr6
      Figure 6Autosomal Dominant Polycystic Kidney Disease Impact Scale (ADPKD-IS), EQ-5D, and 12-Item Short Form Health Survey (SF-12) Cohen d effect size at baseline by chronic kidney disease (CKD) stage. Abbreviations: MCS, Mental Component Summary; PCS, Physical Component Summary.

      Discussion

      At the outset of our research, we were faced with discrepant reports of disease burden in ADPKD and its onset based on literature and physician reports versus anecdotal reports through patient foundations and patients. This discrepancy was also observed in the feedback from clinical experts, who may not be aware of patients’ early concerns due to the very intermittent visit schedules in the earlier disease stages, and patients, who may not bring concerns to physicians because they have adjusted to the disease impact on their lives and experiences with trivialization of these concerns by physicians.
      • Tong A.
      • Rangan G.K.
      • Ruospo M.
      • et al.
      A painful inheritance–patient perspectives on living with polycystic kidney disease: thematic synthesis of qualitative research.
      Lack of patient-centric tools to assess ADPKD-related disease burden has led to a knowledge gap for disease stages, giving the impression that the burden of patients with ADPKD is no different from that of the general population (Fig 4).
      • Tong A.
      • Rangan G.K.
      • Ruospo M.
      • et al.
      A painful inheritance–patient perspectives on living with polycystic kidney disease: thematic synthesis of qualitative research.
      • Miskulin D.C.
      • Abebe K.Z.
      • Chapman A.B.
      • et al.
      Health-related quality of life in patients with autosomal dominant polycystic kidney disease and CKD stages 1-4: a cross-sectional study.
      • Lecardeur L.
      • Joly D.
      [Quality of life in patients with autosomal dominant polycystic kidney disease] in French.
      New initiatives for further understanding of priorities for different stakeholders, including patients, caregivers, physicians, and researchers (such as Standardized Outcomes in Nephrology [SONG]), have been initiated since we started our research, but to date, they focus on other areas in nephrology, and ADPKD-specific outcome measures have yet to emerge.
      The ADPKD-IS is a new tool with support for its reliability (internal consistency and test-retest) and validity (content and construct). The ADPKD-IS is useful for assessing ADPKD-related disease burden across all CKD stages in a cross-sectional cohort, but also tracking disease burden long term.
      The physical domain includes 5 items measuring impact on the ability to perform various activities and 2 items measuring impact of disease-associated pain on daily activities. Identification of 3 distinct types of pain led to retention of the existing pain questions as descriptive of the overall pain concept. However, we also proceeded to develop an additional questionnaire specific to ADPKD-related pain.

      Cheng R, Oberdhan D, Krasa H, Cole JC. Patient-reported pain in autosomal dominant polycystic kidney disease (ADPKD): Concept Stability Between Europe and United States (US) Focus Groups.

      Oberdhan D, Palsgrove A, Cole J, Blais J, Cheng R. Patient experience with pain related to autosomal dominant polycystic kidney disease (ADPKD). Presented at World Congress of Nephrology, March 15, 2015. https://doi.org/10.13140/RG.2.1.2771.4321.

      The fatigue domain assesses 3 specific features of ADPKD-associated fatigue: general fatigue, tiredness while driving, and fatigue after a good night’s sleep. Items within this domain exhibited higher intradomain correlations than interdomain correlations (Table S3). Therefore, fatigue among patients with ADPKD appears likely to encompass aspects of both emotional and physical burden. The emotional domain assesses the emotional impact of ADPKD via 3 concepts common to many instruments (acceptance, anxiety, and sadness) and a fourth disease-specific item (feeling full before appetite is satisfied).
      The ADPKD-IS covers the entire range of health burdens associated with ADPKD across CKD stages in a single instrument, which is not the case for any other PRO instruments. All items were non-normally distributed, and most patients reported at the lower (less affected) end of the range, consistent with the natural history progression of the disease, for which hyperfiltration can compensate for the early loss of kidney tissue, leading to little change in kidney function until decades after birth.
      • Torres V.E.
      • Harris P.C.
      • Pirson Y.
      Autosomal dominant polycystic kidney disease.
      • Grantham J.J.
      Autosomal dominant polycystic kidney disease.
      Individual items also showed limited ceiling effects, indicating the ability of the ADPKD-IS to differentiate between health burdens as patients progress to later stages of the disease, in which HRQoL is more dramatically affected. Consistent with its predicted ability to assess HRQoL across the entire disease spectrum, the ADPKD-IS can show differentiation between disease stages with more sensitivity than general instruments.
      Use of properly developed PRO instruments is a key element of drug development programs using patient-focused end points and for characterization of disease-specific burden with increasing importance given the 21st Century Cures Act requirements. We also see the ADPKD-IS as a tool for researchers and health care providers to better understand ADPKD-specific patient burden with potential use as a patient management tool in clinical practice. Access to the full US-English ADPKD-IS questionnaire, its user manual, and other language versions are available via Mapi Research Trust at https://eprovide.mapi-trust.org/.
      One consideration to note is that there are a variety of techniques to evaluate responsiveness. As the instrument gains use, additional responsiveness evaluations should be performed.

      Acknowledgements

      The authors thank the patients and clinical experts who participated in the development of the ADPKD-IS and the US PKD Foundation and Tess Harris from the UK PKD Charity for support.
      Peer Review: Received December 19, 2016. Evaluated by 2 external peer reviewers and an external methods reviewer, with editorial input from an Associate Editor and the Editor-in-Chief. Accepted in revised form August 20, 2017.

      Supplementary Material

      References

        • Torres V.E.
        • Harris P.C.
        • Pirson Y.
        Autosomal dominant polycystic kidney disease.
        Lancet. 2007; 369: 1287-1301
        • Grantham J.J.
        Autosomal dominant polycystic kidney disease.
        N Engl J Med. 2008; 359: 1477-1485
        • Grantham J.J.
        • Mulamalla S.
        • Swenson-Fields K.I.
        Why kidneys fail in autosomal dominant polycystic kidney disease.
        Nat Rev Nephrol. 2011; 7: 556-566
        • Gabow P.A.
        Gabow PA 1996 PKD Chapter 13 Natural History ADPKD.pdf.
        in: Watson M.L. Torres V.E. Polycystic Kidney Disease. Oxford University Press, New York, NY1996: 333-355
        • Torres V.E.
        • Harris P.C.
        Autosomal dominant polycystic kidney disease: the last 3 years.
        Kidney Int. 2009; 76: 149-168
        • Halvorson C.R.
        • Bremmer M.S.
        • Jacobs S.C.
        Polycystic kidney disease: inheritance, pathophysiology, prognosis, and treatment.
        Int J Nephrol Renovasc Dis. 2010; 3: 69-83
        • Lentine K.L.
        • Xiao H.
        • Machnicki G.
        • Gheorghian A.
        • Schnitzler M.A.
        Renal function and healthcare costs in patients with polycystic kidney disease.
        Clin J Am Soc Nephrol. 2010; 5: 1471-1479
      1. National Kidney and Urologic Diseases Information Clearinghouse. Polycystic Kidney Disease. Bethesda, MD; 2007. https://www.niddk.nih.gov/-/media/FD324862F39546E88D22860766E3A9EA.ashx. Accessed September 23, 2016.

        • Tong A.
        • Rangan G.K.
        • Ruospo M.
        • et al.
        A painful inheritance–patient perspectives on living with polycystic kidney disease: thematic synthesis of qualitative research.
        Nephrol Dial Transplant. 2015; 30: 790-800
        • de Barros B.P.
        • Nishiura J.L.
        • Heilberg I.P.
        • Kirsztajn G.M.
        Ansiedade, depressão e qualidade de vida em pacientes com glomerulonefrite familiar ou doença renal policística autossômica dominante.
        J Bras Nefrol. 2011; 33: 120-128
        • Heiwe S.
        • Bjuke M.
        “An evil heritage”: interview study of pain and autosomal dominant polycystic kidney disease.
        Pain Manag Nurs. 2009; 10: 134-141
        • Miskulin D.C.
        • Abebe K.Z.
        • Chapman A.B.
        • et al.
        Health-related quality of life in patients with autosomal dominant polycystic kidney disease and CKD stages 1-4: a cross-sectional study.
        Am J Kidney Dis. 2014; 63: 214-226
        • Rizk D.
        • Jurkovitz C.
        • Veledar E.
        • et al.
        Quality of life in autosomal dominant polycystic kidney disease patients not yet on dialysis.
        Clin J Am Soc Nephrol. 2009; 4: 560-566
      2. Maruish M. User’s Manual for the SF-12v2 Health Survey. 3rd ed. QualityMetric Inc, Lincoln, RI2012
        • Hays R.D.
        • Kallich J.D.
        • Mapes D.L.
        • Coons S.J.
        • Carter W.B.
        Development of the Kidney Disease Quality of Life (KDQOLTM) instrument.
        Qual Life Res. 1994; 3: 329-338
        • Cleeland C.S.
        • Ryan K.M.
        Pain assessment: global use of the Brief Pain Inventory.
        Ann Acad Med Singapore. 1994; 23: 129-138
        • Davison S.N.
        Pain in hemodialysis patients: prevalence, cause, severity, and management.
        Am J Kidney Dis. 2003; 42: 1239-1247
        • Gamondi C.
        • Galli N.
        • Schönholzer C.
        • et al.
        Frequency and severity of pain and symptom distress among patients with chronic kidney disease receiving dialysis.
        Swiss Med Wkly. 2013; (February 22:w13750)
        • Lecardeur L.
        • Joly D.
        [Quality of life in patients with autosomal dominant polycystic kidney disease] in French.
        Néphrol Thér. 2017; 13: 505-510
      3. Food and Drug Administration. Guidance for Industry Use in Medical Product Development to Support Labeling Claims Guidance for Industry. 2009.

        • Patrick D.L.
        • Burke L.B.
        • Gwaltney C.J.
        • et al.
        Content validity—establishing and reporting the evidence in newly developed patient-reported outcomes (PRO) instruments for medical product evaluation: ISPOR PRO Good Research Practices Task Force Report: part 2—assessing respondent understanding.
        Value Health. 2011; 14: 978-988
        • Patrick D.L.
        • Burke L.B.
        • Gwaltney C.J.
        • et al.
        Content validity—establishing and reporting the evidence in newly developed patient-reported outcomes (pro) instruments for medical product evaluation: ISPOR PRO Good Research Practices Task Force Report: part 1—eliciting concepts for a new PRO instrument.
        Value Health. 2011; 14: 967-977
        • Cronbach L.J.
        • Meehl P.E.
        Construct validity in psychological tests.
        Psychol Bull. 1955; 52: 281-302
        • Keppel G.
        Design and Analysis: A Researcher’s Handbook.
        3rd Ed. Prentice-Hall Inc, Upper Saddle River, NJ1991
        • Martin B.
        • Pathak D.
        • Sharfman M.
        Validity and reliability of the Migraine-Specific Quality of Life Questionnaire (MSQ version 2.1).
        J Head Face Pain. 2000; (Accessed September 23, 2016)
        • Cronbach L.
        Coefficient alpha and the internal structure of tests.
        Psychometrika. 1951; 16: 297-334
        • Nunnally J.C.
        • Bernstein I.H.
        Psychometric Theory.
        3rd ed. McGraw-Hill, New York, NY1994 (Accessed September 26, 2016.)
        • Anastasi A.
        • Urbina S.
        Psychological Testing.
        Prentice Hall, Upper Saddle River, NJ1997
        • Duhachek A.
        • Iacobucci D.
        Alpha’s standard error (ASE): an accurate and precise confidence interval estimate.
        J Appl Psychol. 2004; 89: 792-808
        • Clark L.A.
        • Watson D.
        Constructing validity: basic issues in objective scale development.
        Psychol Assess. 1995; 7: 309-319
        • Cohen J.
        Statistical Power Analysis for the Behavioral Sciences.
        2nd ed. L. Erlbaum Associates, Hillsdale, NJ1988
      4. Fayers P.M. Machin D. Quality of Life. John Wiley & Sons, Ltd, Oxford, UK2015
        • Juniper E.F.
        • Guyatt G.H.
        • Willan A.
        • Griffith L.E.
        Determining a minimal important change in a disease-specific Quality of Life Questionnaire.
        J Clin Epidemiol. 1994; 47: 81-87
        • Nunnally J.
        • Bernstein I.
        • Berge J.
        Psychometric Theory. 1967; (Accessed September 23, 2016)
        • Oberdhan D.
        • Krasa H.B.
        • Schaefer C.
        • Cole J.C.
        Health-related quality of life (HRQOL) measures in autosomal dominant polycystic kidney disease (ADPKD).
        Value Health. 2015; 18 ([abstract]): A512
      5. Cheng R, Oberdhan D, Krasa H, Cole JC. Patient-reported pain in autosomal dominant polycystic kidney disease (ADPKD): Concept Stability Between Europe and United States (US) Focus Groups.

      6. Oberdhan D, Palsgrove A, Cole J, Blais J, Cheng R. Patient experience with pain related to autosomal dominant polycystic kidney disease (ADPKD). Presented at World Congress of Nephrology, March 15, 2015. https://doi.org/10.13140/RG.2.1.2771.4321.