On October 23, 2017, the LURN Steering Committee convened a panel of experts from its membership. The expert review panel was comprised of 24 participating investigators in the LURN Network. These experts included 10 urologists, 5 urogynecologists, 4 patient reported outcome measurement specialists, 3 quantitative specialists, a nephrologist, and a transplant surgeon. Of the 24 participants, 10 were women. Only the LUTS clinicians on the panel (7 women; 8 men) participated in the discussion and decision-making regarding the composition of the brief clinical form.
We employed a modified Delphi approach to facilitating group discussion and consensus-based decision-making. This four-hour meeting had two objectives: 1) To develop an outcome tool with subscales derived from the CASUS questionnaire; and 2) To develop a brief clinical form that improves upon existing measures of clinical symptom reporting, derived from the set of questions selected under the first objective. The first objective included reviewing results of factor analyses and making final decisions on the naming and item content of subscales (Incontinence, Urgency, Voiding Difficulty, Bladder Pain, Nocturia), and selecting additional clinically-relevant questions, resulting in the LURN SI-29 as reported elsewhere.9 The second objective – the brief clinical form – is the subject of this manuscript.
As reported elsewhere, the first half of this meeting, with discussion and input from the entire panel of LURN investigators, resulted in the formation of the LURN SI-29, comprised of five multi-item subscales and nine additional questions not included in any of those subscales. Optional item sets were proffered, discussed, and decided upon through a consensus-building discussion. The resulting composition was unanimously endorsed. Upon completion of this 29-item set, the clinicians on the panel continued discussion in the second half of the meeting, with the instruction of selecting one to two items from each subscale and a finite number of items from among the miscellaneous nine items not included in one of the multi-item subscales. Discussion ensued, scale-by-scale, until consensus was reached on which item (or items) from each set should be included on the clinical form. The focus of discussion was on selecting those items that were regarded as actionable or at least indicative of the need for further clinical interrogation. All item selection decisions were made without member dissent and were thereby assumed to be unanimous.
The 10 items (including frequency, nocturia, urgency, incontinence, bladder pain, voiding, and post-micturition symptoms) were summed to form a 10-item LURN-Symptom Index (LURN SI-10), with score range from 0 to 38. We used all available response data from the 12-month assessment to evaluate the validity of the brief clinical form. Specifically, we compared the LURN SI-10 total score to participant responses on the general screening questions regarding bladder and urinary function described above. Medians and interquartile ranges (IQR) for LURN SI-10 scores were reported by response level on the general screening questions, and relationships between LURN SI-10 scores and responses to the screening questions were examined using proportional odds models. The LURN SI-10 score was then compared to the AUA-SI and the UDI-6 using Pearson correlations and ROC curves. ROC curves were plotted to illustrate the diagnostic ability of various LURN-SI discrimination thresholds as binary classifiers of case definitions defined by the AUA-SI, UDI-6, and the four global screening items from CASUS.
We employed a modified Delphi approach to facilitating group discussion and consensus-based decision-making. This four-hour meeting had two objectives: 1) To develop an outcome tool with subscales derived from the CASUS questionnaire; and 2) To develop a brief clinical form that improves upon existing measures of clinical symptom reporting, derived from the set of questions selected under the first objective. The first objective included reviewing results of factor analyses and making final decisions on the naming and item content of subscales (Incontinence, Urgency, Voiding Difficulty, Bladder Pain, Nocturia), and selecting additional clinically-relevant questions, resulting in the LURN SI-29 as reported elsewhere.9 The second objective – the brief clinical form – is the subject of this manuscript.
As reported elsewhere, the first half of this meeting, with discussion and input from the entire panel of LURN investigators, resulted in the formation of the LURN SI-29, comprised of five multi-item subscales and nine additional questions not included in any of those subscales. Optional item sets were proffered, discussed, and decided upon through a consensus-building discussion. The resulting composition was unanimously endorsed. Upon completion of this 29-item set, the clinicians on the panel continued discussion in the second half of the meeting, with the instruction of selecting one to two items from each subscale and a finite number of items from among the miscellaneous nine items not included in one of the multi-item subscales. Discussion ensued, scale-by-scale, until consensus was reached on which item (or items) from each set should be included on the clinical form. The focus of discussion was on selecting those items that were regarded as actionable or at least indicative of the need for further clinical interrogation. All item selection decisions were made without member dissent and were thereby assumed to be unanimous.
The 10 items (including frequency, nocturia, urgency, incontinence, bladder pain, voiding, and post-micturition symptoms) were summed to form a 10-item LURN-Symptom Index (LURN SI-10), with score range from 0 to 38. We used all available response data from the 12-month assessment to evaluate the validity of the brief clinical form. Specifically, we compared the LURN SI-10 total score to participant responses on the general screening questions regarding bladder and urinary function described above. Medians and interquartile ranges (IQR) for LURN SI-10 scores were reported by response level on the general screening questions, and relationships between LURN SI-10 scores and responses to the screening questions were examined using proportional odds models. The LURN SI-10 score was then compared to the AUA-SI and the UDI-6 using Pearson correlations and ROC curves. ROC curves were plotted to illustrate the diagnostic ability of various LURN-SI discrimination thresholds as binary classifiers of case definitions defined by the AUA-SI, UDI-6, and the four global screening items from CASUS.
