Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Q12
Q1
Clinical Gastroenterology and Hepatology 2015;-:-–-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
Development and Validation of an Inflammatory BowelDiseases Monitoring Index for Use With MobileHealth Technologies
68
69
70
71
72
73
74
75
76
Welmoed K. Van Deen,*,‡ A. E. van der Meulen-de Jong,‡ N. K. Parekh,§ E. Kane,*A. Zand,* C. A. DiNicola,* L. Hall,* E. K. Inserra,* J. M. Choi,* Ch. Y. Ha,* E. Esrailian,*M. G. H. van Oijen,*,jj and D. W. Hommes*
*UCLA Center for Inflammatory Bowel Diseases, Melvin and Bren Simon Digestive Diseases Center, David Geffen Schoolof Medicine, University of California, Los Angeles, Los Angeles, California; ‡Department of Gastroenterology andHepatology, Leiden University Medical Center, Leiden, the Netherlands; §Division of Gastroenterology andHepatology, University of California, Irvine, Irvine, California; and jjDepartment of Medical Oncology, Academic MedicalCenter, Amsterdam, the Netherlands
77
78
BACKGROUND & AIMS:Abbreviations used in this papedisease; CDAI, Crohn’s diseasvisual analogue scale; HBI, HaIBD, inflammatory bowel diseCenter; mHealth, mobile healthpredictive value; pMayo, partivalue; PROs, patient-reportedacteristics; SCCAI, simple clin
F
79
80
81
82
Mobile health technologies are advancing rapidly as smartphone use increases. Patients withinflammatory bowel disease (IBD) might be managed remotely through smartphone applica-tions, but no tools are yet available. We tested the ability of an IBD monitoring tool, which canbe used with mobile technologies, to assess disease activity in patients with Crohn’s disease(CD) or ulcerative colitis (UC).
83
84
METHODS:85
86
87
88
89
90
We performed a prospective observational study to develop and validate a mobile health indexfor CD and UC, which monitors IBD disease activity using patient-reported outcomes. Wecollected data from disease-specific questionnaires completed by 110 patients with CD and 109with UC who visited the University of California, Los Angeles, Center for IBD from May 2013through January 2014. Patient-reported outcomes were compared with clinical disease activityindex scores to identify factors associated with disease activity. Index scores were validated in301 patients with CD and 265 with UC who visited 3 tertiary IBD referral centers (in Californiaor Europe) from April 2014 through March 2015.
91
92
RESULTS: 9394
95
96
97
98
99
We assessed activity of CD based on liquid stool frequency, abdominal pain, patient well-being,and patient-assessed disease control, and activity of UC based on stool frequency, abdominalpain, rectal bleeding, and patient-assessed disease control. The indices identified clinical dis-ease activity with area under the receiver operating characteristic curve values of 0.90 in pa-tients with CD and 0.91 in patients with UC. They identified endoscopic activity with area underthe receiver operating characteristic values of 0.63 in patients with CD and 0.82 in patients withUC. Both scoring systems responded to changes in disease activity (P < .003). The intraclasscorrelation coefficient for test-retest reliability was 0.94 for CD and for UC.
100
101
CONCLUSIONS:102
103
We developed and validated a scoring system to monitor disease activity in patients with CDand UC that can be used with mobile technologies. The indices identified clinical disease activitywith area under the receiver operating characteristic curve values of 0.9 or higher in patientswith CD or UC, and endoscopic activity in patients with UC but not CD.
104
105
Keywords: Patient-Reported Outcomes; Telemedicine; Severity; Management.106
107
r: AUC, area under the curve; CD, Crohn’se activity index; DA-VAS, disease activityrvey Bradshaw index; Hgb, hemoglobin;ase; LUMC, Leiden University Medical; mHI, mobile health index; NPV, negativeal Mayo score; PPV, positive predictiveoutcomes; ROC, receiver operating char-ical colitis activity index; SES-CD, simple
endoscopic index for Crohn’s disease; UC, ulcerative colitis; UCI,University of California, Irvine; UCLA, University of California, Los Angeles;VAS, visual analogue scale.
© 2015 by the AGA Institute1542-3565/$36.00
http://dx.doi.org/10.1016/j.cgh.2015.10.035
LA 5.4.0 DTD � YJCGH54544_proof � 11 December 2015 � 12:31 pm � ce CJ
108
109
110
111
112
113
114
115
116
Q5 Q6
Q7
Q8
2 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
The shift from symptom-oriented to prevention-oriented care delivery has accelerated the devel-
opment of mobile health (mHealth) technologies and isthought to radically transform health care delivery.1
Smartphone adoption is increasing rapidly, with 64%of Americans using smartphones in 2014 of which 62%used their telephones to look up health information.2
Many health apps are available, most of which providehealth information or support data collection.3 For pa-tients with inflammatory bowel diseases (IBD), appsare available that assist in tracking symptoms, loggingmeals, and managing medications.4,5 These apps cancreate reports for providers but do not allow for real-time interactions between patient and provider.
Self-monitoring and self-management for chronicdiseases is widely practiced in diabetes care6 and anti-coagulation therapy.7 Additionally, e-technologies forsymptom reporting between patients and providers areincreasingly used in chronic diseases.8 Several systemsfor online symptom reporting and disease managementhave been developed for IBD. In the Danish Constant-Care web system, patients with ulcerative colitis (UC)filled out a clinical symptom score and logged fecal cal-protectin levels weekly; based on these scores the sys-tem made real-time recommendations for adjustingmesalamine dosing. This approach was shown toempower patients and decrease relapse duration.9,10
Similarly, individualization of infliximab dosing in pa-tients with Crohn’s disease (CD) was reported to bepractical and feasible.11 A study evaluating another hometele-management system in UC (UC-HAT) did not showsignificant differences in disease activity and quality oflife between users and control subjects, and more thanone-third of the patients discontinued participation.12 Anongoing multicenter randomized controlled trial istesting the use of a mobile tele-management system us-ing text messaging in IBD. This system sends personal-ized alerts and educational texts, and assesses symptomsand side effects, based on which treatment can bemodified.13
Patient-reported outcomes (PROs) are increasinglyused to evaluate health status, and the importance ofPROs in outcome measurement, symptom management,and quality improvement efforts is increasingly recog-nized.14 Furthermore, the use of PROs as primaryoutcome measures for evaluating effectiveness of IBDinterventions is progressively supported by the Food andDrug Administration.15 Therefore, PROs are promisingfor use in mHealth apps. An example is the Health-PROMISE app, which tracks patient-reported quality oflife scores in patients with IBD and provides decisionsupport to physicians.16 However, accurate e-monitoringtools for disease activity in IBD are yet to be developed.Previous efforts have aimed to develop PRO question-naires by adjusting existing scores.17–19 We aimed toidentify the most optimal PRO score to use on an IBDdisease-monitoring app. The best PROs were selected
FLA 5.4.0 DTD � YJCGH54544_proof � 1
from an exhaustive list of PROs in a prospective cohort ofpatients with IBD. Subsequently, the developed scoreswere tested prospectively in an independent cohort at 3independent IBD centers.
Methods
Design
We performed a prospective, observational study thataimed to develop and validate an mHealth index (mHI)for CD and UC that accurately monitors IBD diseaseactivity using PROs. The study consisted of 2 phases: adevelopment phase and a validation phase. During thedevelopment phase the mHIs were developed usingcollected PROs and clinical disease activity indices.During the validation phase the developed mHIs werevalidated in an independent cohort.
Population
Development phase. Patients with IBD were identi-fied during clinic visits between May 2013 and January2014 at the University of California, Los Angeles (UCLA)Center for IBD. Patients with esophageal or anal CDinvolvement alone, patients with a pouch or stoma, andpregnant women were excluded. Eligible patients filledout disease-specific questionnaires assessing PROs ofmost common clinical disease activity indices: partialMayo (pMayo), simple clinical colitis activity index(SCCAI), and modified Truelove and Witts index for UC;Harvey Bradshaw index (HBI) and CD activity index(CDAI) including a 7-day diary before the visit(Supplementary Figure 1) for CD. Additionally, patientswere asked to assess their symptoms and perceiveddisease activity using visual analogue scales (VAS). ThePROs were categorized into 10 domains: stool fre-quency, abdominal pain, general well-being, urgency,stool consistency, rectal bleeding, fever, anorexia,nausea/vomiting, and perceived disease activity(Table 1).
During clinic visits, vital signs were measured andphysicians collected the physician-reported outcomesrequired for the clinical disease activity indices(Table 1). Hemoglobin (Hgb), hematocrit, white bloodcell count, platelets, albumin, C-reactive protein, anderythrocyte sedimentation rate were requested.Furthermore, stool testing for calprotectin was reques-ted either at the patient’s preferred laboratory or usinga free stool kit (Genova Diagnostics, Asheville, NC)picked up at the patient’s home. A dedicated studynurse (E.K.I.) followed up with patients via telephone ore-mail to ensure laboratory and stool tests wereperformed.
Validation phase. Eligible patients with IBD wereidentified during clinic and endoscopy visits between
1 December 2015 � 12:31 pm � ce CJ
232
Q9
Table 1. Collected PROs and Clinical Markers in CD and Patients With UC
Domain Question CD UC
1. Stool frequency Number of liquid/very soft stools for each of the last 7 days XHow many stools did you have yesterday during the day? X XHow many stools did you have yesterday during the night? X XHow many stools do you have normally during the day? X XHow many stools do you have normally during the night? X X
2. Abdominal pain Abdominal pain for each of the last 7 days (no/mild/moderate/severe) XAbdominal pain (no/mild/moderate/severe) XAbdominal pain (no abdominal pain/with bowel actions/continuous) X XRate your abdominal pain on a scale from 0 to 10 (VAS) X X
3. General well-being General well-being for each of the last 7 days (very well/slightly below par/poor/very poor/terrible)
X
General well-being (very well/slightly below par/poor/very poor/terrible) XGeneral well-being (perfect/very good/good/average/poor/terrible) XWell-being (no impairment/impaired, but able to continue activities/activities reduced/unable
to work)X X
Rate your well-being on a scale from 0 to 10 (VAS) X X4. Urgency Urgency of defecation (no urgency/hurry/immediate/incontinence) X X5. Stool consistency Stool consistency (normal or variably normal/semiformed/liquid) X X
Do you take opiates or lomotil/imodium for diarrhea? X XHow often do you take antidiarrheals? (0–10 VAS) X X
6. Rectal bleeding What % of bowel movements contains visible blood? (none/less than 50%/50% or more/blood alone)
X X
Amount of blood in stool (none/trace/occasionally frank [bright red]/usually frank) X XHow often do you experience rectal bleeding? (0–10 VAS) X X
7. Fever Fever on each of the last 7 days X8. Anorexia Loss of appetite (yes/no) X X9. Nausea/vomiting Nausea and/or vomiting (yes/no) X X10. Disease activity How well do you feel your disease is under control? (0–10 VAS) X X11. Clinical markers Temperature X X
Weight and height XPulse XAbdominal tenderness XAbdominal mass XExtraintestinal manifestations X XPhysician global assessment of disease activity X XHgb, hematocrit, white blood cell, platelets, albumin, C-reactive protein, erythrocyte
sedimentation rate (blood) and calprotectin (stool)X X
- 2015 mHealth Index for IBD Monitoring 3
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
April 2014 and March 2015 in 3 tertiary IBD referralcenters (UCLA; University of California, Irvine [UCI]; andLeiden University Medical Center [LUMC], theNetherlands). Patients who participated in the devel-opment phase of the study were excluded. For patientswith CD, the developed mHI-CD and HBI werecompleted during clinic visits; during endoscopic visits,the simple endoscopic score for CD (SES-CD) was addi-tionally completed. For patients with UC, the mHI-UCand pMayo were collected during clinic visits, and theMayo endoscopic subscore was additionally obtainedduring endoscopic visits. Patients at LUMC completed aDutch version of the mHI-CD and mHI-UC. After trans-lation to Dutch, the questionnaires were translated backto English by an independent translator; the Dutchquestionnaire was then revised and retested. To assesssensitivity of the mHI to detect changes in clinical dis-ease activity, a subset of patients was included a secondtime during scheduled follow-up visits. To assess test-test reliability, a subset of UCLA patients was asked to
FLA 5.4.0 DTD � YJCGH54544_proof � 1
complete a second questionnaire at home after theirclinic visit.
Definitions
For CD, clinical disease activity was defined as an HBI>4 or a CDAI >150. A change of �3 in HBI wasconsidered a clinically relevant change.20 Endoscopicdisease activity was defined as an SES-CD >3. For UC,clinical disease activity was defined as a pMayo >2, a PTI>3, or an SCCAI >2. A change of �3 in the pMayo wasconsidered a clinically relevant change.18 Endoscopicdisease activity was defined as a Mayo endoscopic sub-score >1.
Ethical Considerations
All patients consented to participate in this study.This study was approved by the institutional review
1 December 2015 � 12:31 pm � ce CJ
4 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
boards of participating centers under the followingprotocol numbers: UCLA, IRB#13-000402; UCI, HS#2014-1231; and LUMC, P14.158.
Statistical Analysis
Descriptive statistics were used for clinical charac-teristics and demographic information. Numeric valuesare presented as mean and standard deviation or medianand range. SAS version 9.2 (SAS Institute, Cary, NC) wasused for statistical analyses.
Development phase. Univariate logistic regressionwas performed using disease activity (HBI >4 for CD orpMayo >2 for UC) as the dependent variable and thePROs as independent variables. For each of the PROs,different cutoffs were used, which roughly created linearassociations between the groups and the chance of activedisease. Because different PROs represented the samedomain (Table 1), the variables with the highest Waldchi-square value for predicting clinical disease activitywere selected within each domain for inclusion in themultivariate logistic regression models. If 2 variableswithin the same domain had a similar predictive value(difference between Wald chi-square values <0.5), bothwere tried in separate models unless the question typewas less preferable. Because of usability on a mobileapplication VAS, yes/no, or numeric questions werepreferred over categorical questions; within those,questions with less response options were preferred.Variables with a P value >.1 in the univariate analysiswere omitted from subsequent analyses. Stepwise for-ward multivariate logistic regression was performedwith clinical disease activity as the dependent variableand the selected PROs as independent variables. A sig-nificance level of P < .1 was required for entry in themodel and a significance level of P < .1 to stay in themodel. Several models were performed using differentclinical disease activity indices to define clinical diseaseactivity as the dependent variable.
Composite scores were created using the regressioncoefficients of independent predictors in the multivariatemodel. Spearman correlation coefficients were estimatedbetween the newly developed mHIs and clinical diseaseactivity indices. Receiver operating characteristic (ROC)curves were used to assess the capability of the mHI todiscriminate active versus nonactive disease usingdifferent clinical disease activity indices, and the areasunder the curves (AUC) were calculated. The compositescore with overall highest AUC using different goldstandards was selected.
Because the main aim of the developed score was toidentify patients at risk for active disease, we defined theoptimal cutoff for disease activity as a negative predic-tive value (NPV) of �95% and a sensitivity of �85%while maintaining maximum specificity. The overallprevalence of active disease was estimated at 22% basedon cross-sectional cohort data from UCLA.
FLA 5.4.0 DTD � YJCGH54544_proof � 1
Validation phase. To validate the mHI against clinicaland endoscopic disease activity indices, the mHI-UC wascompared with the pMayo and the endoscopic subscoreof the Mayo; the mHI-CD was compared with the HBI andSES-CD. Spearman correlation coefficients between thescores were calculated and ROC curves to assess theability to predict clinical and endoscopic disease activitywere generated. To assess sensitivity to change, wecompared patients who clinically improved, remainedstable, and deteriorated. A Kruskal-Wallis test was usedto compare groups. Test-retest reliability was assessedby the intraclass correlation coefficient. The performanceof the VAS for patient-reported disease activity (DA-VAS)as single predictor for clinical and endoscopic diseaseactivity also was assessed.
Results
Development Phase
In total, 219 patients (110 CD and 109 UC) wereincluded in the development phase of the study(Figure 1A, Table 2). In 108 out of 110 patients with CDthe HBI was calculated, whereas the CDAI could only becalculated in 93 out of 110. The pMayo, SCCAI, andmodified Truelove and Witts index were calculated in allpatients with UC. Complete laboratory and stool testswere obtained from only 48% of patients. Despiteintensive follow-up by a dedicated research nurse(E.K.I.), 39% of patients did not perform stool testing and13% did not have laboratory values drawn. Additionally,14% of patients had blood drawn, but laboratory setswere incomplete because of protocol deviations.
Univariate logistic regression was performed forPROs and blood and stool tests (Supplementary Tables 1and 2). Stool frequency, abdominal pain, general well-being, urgency, and patient-reported disease activitywere all strong predictors for clinical disease activity inboth CD and UC (P < .0001). Incontinence was onlypresent in 3% of patients and did not predict diseaseactivity in either CD (P ¼ .54) or UC (P ¼ .99). In CD theuse of antidiarrheals was predictive for disease activity(P ¼ .019) but not in UC (P ¼ .96), whereas the VASassessing frequency of antidiarrheal use was a predictorfor neither CD (P ¼ 1.00) nor UC (P ¼ .26). Rectalbleeding was a predictor for disease activity in both UC(P < .0001) and CD (P ¼ .019). Anorexia was predictivein both CD (P ¼ .019) and UC (P ¼ .0025), whereasnausea and vomiting predicted only CD disease activity(P ¼ .035) and not UC disease activity (P ¼ .14). High C-reactive protein (P ¼ .0009), high erythrocyte sedimen-tation rate (P ¼ .0022), low Hgb (P ¼ .022), and lowalbumin (0.034) were predictors for clinical disease ac-tivity in CD. High calprotectin was not a significant pre-dictor for CD disease activity (P ¼ .054), althoughcalprotectin as continuous variable had predictive value
1 December 2015 � 12:31 pm � ce CJ
464
Figure 1. Inclusion flow-charts for developmentphase (A) and validationphase (B).
- 2015 mHealth Index for IBD Monitoring 5
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
(P ¼ .011). Low platelets (P ¼ .98), low hematocrit (P ¼.28), and high white blood cell (P ¼ .11) were not pre-dictive in CD (Supplementary Table 1). In UC highC-reactive protein (P ¼ .0067), high calprotectin (P ¼.022), high white blood cell (0.013), high erythrocytesedimentation rate (P ¼ .028), and low hematocrit (P ¼.0047) were all predictive for clinical disease activity.Low albumin (P ¼ .98) was not predictive for clinicalactivity in UC, although albumin as continuous variablewas (P ¼ .02). Low platelets (P ¼ .99) and low Hgb (P ¼.13) were not predictive for disease activity in UC,whereas Hgb as continuous variable (P ¼ .032) was(Supplementary Table 2).
FLA 5.4.0 DTD � YJCGH54544_proof � 1
The most representative PROs were selected for in-clusion in the multivariate regression model(Supplementary Tables 1 and 2). Because of lowcompletion rates of laboratory testing despite intensivefollow-up, laboratory tests were initially excluded fromthe multivariate analysis and score development. In CD, 4composite scores were developed (SupplementaryTable 3); in UC, 11 composite scores were evaluated(Supplementary Table 4). Addition of laboratory vari-ables to the selected models decreased the sample sizeand therefore the power of the regression models;addition of the laboratory variables to the model did notresult in inclusion of these variables in the models,
1 December 2015 � 12:31 pm � ce CJ
580
Table 2. Demographics of Included Patients in Development and Validation Phase of the Study
Development phase Validation phase
CD UC CD UC
N 110 109 301 265Median age, median (range) 33 (19–79) 35 (18–81) 33 (18–75) 42 (18–86)Male, n (%) 56 (51) 57 (52) 144 (48) 132 (50)a
Smoking, n (%) 9 (8) 7 (6)a 19 (6)b 12 (5)c
Age at diagnosis, median (range) 24 (8–68) 28 (10–81) 25 (8–66) 29 (2–76)Disease duration, median (range) 8 (0–46) 6 (0–52) 8 (0–52) 7 (0–59)Surgical history, n (%) 48 (44) 1 (1) 132 (44)d 13 (5)d
CD fistulizing disease, n (%) 5 (5)a — 59 (20)a —
Active disease (HBI >4 /pMayo >2), n (%) 32 (30)a 37 (34) 82 (27) 82 (31)
an ¼ 1 unknown.bn ¼ 8 unknown.cn ¼ 9 unknown.dn ¼ 2 unknown.
6 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
because they did not reach the required significance levelof P < .1 for entry in the model.
For CD the selected composite score (Table 3) had anAUC of >0.95 for predicting clinical disease activity,using both CDAI and HBI as gold standards (0.951 and0.963, respectively). The Spearman correlation co-efficients were 0.837 and 0.830, respectively(Supplementary Table 3). The optimal cutoff for themHI-CD to predict clinical disease activity was set at�5.5, resulting in an NPV of 96%, positive predictivevalue (PPV) of 63%, sensitivity of 88%, and specificityof 85%. For UC the selected composite score (Table 3)had an AUC of >0.91 to predict disease activity usingpMayo, SCCAI, and modified Truelove and Witts indexas gold standards (0.960, 0.915, and 0.913, respec-tively). The Spearman correlation coefficients were0.820, 0.832, and 0.790, respectively (SupplementaryTable 4). The optimal cutoff for the mHI-UC to predictclinical disease activity was set at �4.99, resulting in anNPV of 97%, PPV of 72%, sensitivity of 89%, andspecificity of 90%.
Table 3. Calculation of the mHI-CD and mHI-UC
mHI-CD questions Answer Score
Number of liquid/very soft stools/day 0 0.0000 How mhave1–2 1.6983
�3 3.3966Abdominal pain No 0.0000 Rate yo
scaleYes 2.3868
Rate your well-being on a scale from0 to 10 (0 ¼ worst, 10 ¼ best)
8–10 0.0000 How of(0 ¼4–7 2.1336
0–3 4.2672How well do you feel your disease is
under control (0 ¼ no disease activity,10 ¼ worst disease activity)
0–2 0.0000 How w(0 ¼3–6 2.1175
7–10 4.2350Total score (SUM) Total s
FLA 5.4.0 DTD � YJCGH54544_proof � 1
Validation Phase
A total of 301 patients with CD (UCLA, n ¼ 127; UCI,n ¼ 82; LUMC, n ¼ 92) and 265 patients with UC (UCLA,n ¼ 119; UCI, n ¼ 67; LUMC, n ¼ 79) were analyzed inthe validation phase (Figure 1B, Table 2). For CD theSpearman correlation coefficient was 0.75 (P < .0001)between HBI and mHI-CD, the AUC of the ROC for pre-dicting clinical disease activity was 0.90, and a sensitivityof 94% and specificity of 67% were achieved using acutoff of �5.5 (Figures 2A and 2B). To achieve theoptimal NPV of 95% and sensitivity of 85%, the cutoff inthis cohort would be �6.38, which would result in 85%sensitivity, 80% specificity, 95% NPV, and 55% PPV. ThemHI-CD was poorly correlated with the SES-CD (r ¼ .30;P ¼ .0039), with an AUC of 0.63 (Figures 2A and 2C).
For UC the Spearman correlation coefficient was 0.72(P < .0001) between pMayo and mHI-UC, the AUC of theROC for predicting clinical disease activity was 0.91, anda sensitivity of 73% and specificity of 90% specificitywere achieved using a cutoff of �4.99 (Figures 2D and
mHI-UC questions Answer Score
any stools did youyesterday?
�2 0.00003–4 1.4428>4 2.8856
ur abdominal pain on afrom 0 to 10 (0 ¼ none, 10 ¼ worst)
0–2 0.00003–6 1.03927–10 2.0784
ten do you experience rectal bleeding?none, 10 ¼ always)
0–3 0.00004–10 2.2019
ell do you feel your disease is under controlno disease activity, 10 ¼ worst disease activity)
0–2 0.00003–5 1.75576–10 3.5114
core (SUM)
1 December 2015 � 12:31 pm � ce CJ
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
Figure 2. Performance of the developed mHI scores to detect disease activity in CD (A–C) and UC (D–F). ROC curves of themHI to predict clinical and endoscopic disease activity (A and D), scatter plot of the mHI versus clinical disease activity scores(B and E), and endoscopic disease activity scores (C and F).
- 2015 mHealth Index for IBD Monitoring 7
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
2E). The optimal cutoff in this cohort would be �3.2,which would result in 85% sensitivity, 80% specificity,95% NPV, and 55% PPV. The mHI-UC was stronglycorrelated with the Mayo endoscopic subscore (r ¼ .60;P < .0001), with an AUC of 0.82 (Figures 2D and 2F).
Sensitivity to change was assessed in a subset of 50patients with CD and 44 patients with UC. Median timebetween questionnaires was 46 days (range, 2–352) forCD and 57.5 days (range, 3–275) for UC. Four (8%) pa-tients with CD deteriorated, 31 (62%) had stable diseaseactivity, and 15 (30%) improved. Of the patients with UC,5 (11%) deteriorated, 27 (61%) remained stable, and 12(27%) improved. There was a significant difference inmHI between patients who clinically improved, remainedstable, or worsened, with either CD (P ¼ .0030) or UC(P ¼ .0025) (Figures 3A and 3C). Test-retest reliabilitywas assessed in a subset of 40 patients with CD and 37patients with UC. The median time to second question-naire completion was 21 hours (range, 7–36) for the mHI-CD and 23 hours (range, 11–144) for the mHI-UC. Theintraclass correlation coefficient was 0.94 (confidencelimits, 0.89–0.97) for the mHI-CD and 0.94 (confidencelimits, 0.89–0.97) for the mHI-UC (Figures 3B and 3D).
One question in both the mHI-CD and the mHI-UCassessed patient-reported disease activity using a VAS(DA-VAS). In patients with CD (n ¼ 301) the DA-VAS had
FLA 5.4.0 DTD � YJCGH54544_proof � 1
a Spearman correlation of 0.63 (P < .0001) with the HBI,and the AUC for predicting clinical disease activity was0.83 (compared with r ¼ .75 and AUC ¼ 0.90 for the fullmHI-CD). The CD DA-VAS was weakly correlated with theSES-CD (r ¼ .21; P ¼ .040), and had an AUC to predictendoscopic disease activity of 0.59 (compared with r ¼0.30 and AUC ¼ 0.63 for the full mHI-CD). The DA-VASwas not significantly different among patients whodeteriorated, remained stable, or improved (P ¼ .12). Inpatients with UC (n ¼ 265) the DA-VAS had a Spearmancorrelation of 0.67 (P < .0001) with the pMayo, and theAUC for predicting clinical disease activity was 0.86(compared with r ¼ .72 and AUC ¼ 0.91 for the full mHI-UC). The UC DA-VAS was also correlated with theendoscopic component of the Mayo score (r ¼ .55; P <.0001), and had an AUC to predict endoscopic diseaseactivity of 0.79 (compared with r ¼ .60 and AUC ¼ 0.82for the full mHI-UC). A significant difference between theDA-VAS of patients with UC that deteriorated, remainedstable, or improved was observed (P ¼ .0052).
Discussion
We developed 2 questionnaires of 4 items consistingsolely of PROs for remote monitoring of patients with
1 December 2015 � 12:31 pm � ce CJ
Figure 3. Sensitivity tochange (A and C) and test-retest reliability (C and D)of the mHI-CD (A and B)and mHI-UC (C and D).
8 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
IBD, which can be used on mobile technology. Thequestionnaires were validated in a multicenter validationstudy and showed excellent characteristics to monitorclinical disease activity and symptom changes. As previ-ously shown, UC clinical disease activity highly correlateswith endoscopic disease activity, whereas correlationbetween CD symptoms and endoscopic findings is poor.21
Although previous studies have aimed to identifyPROs for disease monitoring either by adjusting existingquestionnaires,16 or by using subcomponents of existingquestionnaires,18,19 we were able to prospectively iden-tify PROs relevant for clinical disease monitoring andvalidate those in an independent cohort. Interestingly,patient-reported disease activity was shown to be anindependent predictor for clinical disease activity inpatients with UC and patients with CD, even after in-clusion of common IBD symptoms, such as stool fre-quency and abdominal pain. Patient-reported diseaseactivity alone had a comparable performance with thecomplete mHI in both CD and UC for detecting clinicaland endoscopic activity, although responsiveness tochanges in disease activity was reduced in particular inpatients with CD.
FLA 5.4.0 DTD � YJCGH54544_proof � 1
A limitation of this study is the potential for recallbias in CDAI calculation. Although 7-day diary formswere sent out in advance, we did not log whether diarieswere filled out daily or by recall. Additionally, in thevalidation cohort, optimal cutoffs of the mHI for detec-tion of disease activity were higher than expected for CDand lower than expected for UC. This might be caused bythe reduction of questionnaire items from >20 PROs tojust 4, or by differences in the patient population. Thevalidation phase of the study most accurately representsthe real-life situation. Therefore, we implemented thecutoff for optimal sensitivity and specificity as observedin the validation cohort in clinical practice.
This study was not primarily designed to evaluatecorrelations between PROs and endoscopic healing. In at-risk patients, clinical assessments remain warranted,which may lead to further endoscopic evaluation. Thistool offers an optimal screening method to monitor andevaluate disease activity in and outside of clinical prac-tice with a high NPV. The mHIs are currently imple-mented in the UCLA eIBD patient app (SupplementaryFigure 2) and automated messages are sent to a nursecoordinator when the mHI indicates disease activity. The
1 December 2015 � 12:31 pm � ce CJ
Q10
- 2015 mHealth Index for IBD Monitoring 9
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
app is currently available to patients treated at the UCLACenter for IBD and can be downloaded in iOS or Androidby searching for “UCLA eIBD.”
The calculation of the mHI is complex. However,simplifying the score would most likely result in loss ofaccuracy. Because the index is meant to be automatedand implemented on digital platforms, we believe thatusing the more complex calculations is justified.
Cloud-based health technologies are predicted torevolutionize care delivery and patient engagement. Pa-tients can participate in their care by signaling meaningfulhealth outcomes during year-round monitoring. Barriersfor more widespread implementation of mHealth in IBDcare include policies affecting reimbursement and regula-tory requirements,22 and privacy and security concerns.1
In summary, we developed the mHI-CD and mHI-UCfor remote monitoring of patients with CD and patientswith UC. The scores are specifically designed for imple-mentation on a mobile application and are currentlyavailable to patients with IBD treated at the UCLA Centerfor IBD. Prospective randomized studies need to assessthe effect of remote monitoring on disease control,quality of life, patient satisfaction, and health care costs.
1010
1011
1012
1013
1014
1015
1016
1017
Supplementary Material
Note: To access the supplementary material accom-panying this article, visit the online version of ClinicalGastroenterology and Hepatology at www.cghjournal.org,and at http://dx.doi.org/10.1016/j.cgh.2015.10.035.
1018
1019
ReferencesQ2
Q3
Q4
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1. Steinhubl SR, Muse ED, Topol EJ. The emerging field of mobilehealth. Sci Transl Med 2015;7:283rv3.
2. The Smartphone Difference. Pew Research Center, April2015. Available at: http://www.pewinternet.org/2015/04/01/us-smartphone-use-in-2015/. Accessed: May 11, 2015.
3. Kuehn BM. Is there an app to solve app overload? JAMA 2015;313:1405–1407.
4. Available at: http://www.medivo.com/gi/.
5. Available at: http://www.ibdetermined.org/Tracker.aspx.Accessed: May 13, 2015.
6. Haas L, Maryniuk M, Beck J, et al. National standards for dia-betes self-management education and support. Diabetes Care2013;36(Suppl 1):S100–108.
7. Heneghan C, Ward A, Perera R, et al. Self-monitoring of oralanticoagulation: systematic review and meta-analysis of indi-vidual patient data. Lancet 2012;379:322–334.
8. Johansen MA, Henriksen E, Horsch A, et al. Electronic symptomreporting between patient and provider for improved health careservice quality: a systematic review of randomized controlledtrials. part 1: state of the art. J Med Internet Res 2012;14:e118.
9. Elkjaer M, Shuhaibar M, Burisch J, et al. E-health empowers pa-tients with ulcerative colitis: a randomised controlled trial of theweb-guided ’Constant-care’ approach. Gut 2010;59:1652–1661.
10. Pedersen N, Thielsen P, Martinsen L, et al. eHealth: individual-ization of mesalazine treatment through a self-managed
FLA 5.4.0 DTD � YJCGH54544_proof � 1
web-based solution in mild-to-moderate ulcerative colitis.Inflamm Bowel Dis 2014;20:2276–2285.
11. Pedersen N, Elkjaer M, Duricova D, et al. eHealth: individuali-sation of infliximab treatment and disease course via a self-managed web-based solution in Crohn’s disease. AlimentPharmacol Ther 2012;36:840–849.
12. Cross RK, Cheevers N, Rustgi A, et al. Randomized, controlledtrial of home telemanagement in patients with ulcerative colitis(UC HAT). Inflamm Bowel Dis 2012;18:1018–1025.
13. Cross RK, Jambaulikar G, Langenberg P, et al. TELEmedicinefor Patients with Inflammatory Bowel Disease (TELE-IBD):design and implementation of randomized clinical trial. ContempClin Trials 2015;42:132–144.
14. Jensen RE, Rothrock NE, DeWitt EM, et al. The role of technicaladvances in the adoption and integration of patient-reportedoutcomes in clinical care. Med Care 2015;53:153–159.
15. Williet N, Sandborn WJ, Peyrin-Biroulet L. Patient-reportedoutcomes as primary end points in clinical trials of inflam-matory bowel disease. Clin Gastroenterol Hepatol 2014;12:1246–1256.e6.
16. Atreja A, Khan S, Rogers JD, et al. Impact of the MobileHealthPROMISE platform on the quality of care and quality oflife in patients with inflammatory bowel disease: study protocolof a pragmatic randomized controlled trial. JMIR Res Protoc2015;4:e23.
17. Evertsz FB, Hoeks CC, Nieuwkerk PT, et al. Development of thepatient Harvey Bradshaw index and a comparison with aclinician-based Harvey Bradshaw index assessment of Crohn’sdisease activity. J Clin Gastroenterol 2013;47:850–856.
18. Lewis JD, Chuai S, Nessel L, et al. Use of the noninvasivecomponents of the Mayo score to assess clinical response inulcerative colitis. Inflamm Bowel Dis 2008;14:1660–1666.
19. Khanna R, Zou G, D’Haens G, et al. A retrospective analysis: thedevelopment of patient reported outcome measures for theassessment of Crohn’s disease activity. Aliment Pharmacol Ther2015;41:77–86.
20. Vermeire S, Schreiber S, Sandborn WJ, et al. Correlation be-tween the Crohn’s disease activity and Harvey-Bradshawindices in assessing Crohn’s disease severity. Clin Gastro-enterol Hepatol 2010;8:357–363.
21. Falvey JD, Hoskin T, Meijer B, et al. Disease activity assessmentin IBD: clinical indices and biomarkers fail to predict endoscopicremission. Inflamm Bowel Dis 2015;21:824–831.
22. Adler-Milstein J, Kvedar J, Bates DW. Telehealth among UShospitals: several factors, including state reimbursement andlicensure policies, influence adoption. Health Aff 2014;33:207–215.
Reprint requestsAddress requests for reprints to: Welmoed K. Van Deen, MD, Center forInflammatory Bowel Diseases, Melvin and Bren Simon Digestive DiseasesCenter, David Geffen School of Medicine, University of California at LosAngeles, 10945 Le Conte Avenue, #2114, Los Angeles, California 90095.e-mail: [email protected]; fax: (310) 206-9906.
Conflicts of interestThe authors disclose no conflicts.
FundingThis study was partly supported by a grant from Genova Diagnostics. GenovaDiagnostics provided stool collection kits and performed fecal calprotectintesting during the study.
1 December 2015 � 12:31 pm � ce CJ
1043
1044
SupplementaryFigure 1. Seven-day diaryform, which was sent topatients to log symptomsduring the 7 days beforethe clinic visit for calcula-tion of the CDAI.
web4C=FPO
9.e1 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
FLA 5.4.0 DTD � YJCGH54544_proof � 11 December 2015 � 12:31 pm � ce CJ
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
web4C=FPO
Supplementary Figure 2. Screenshots of the mHI as implemented in the UCLA eIBD patient app available to patients treatedat the UCLA IBD Center.
FLA 5.4.0 DTD � YJCGH54544_proof � 11 December 2015 � 12:31 pm � ce CJ
- 2015 mHealth Index for IBD Monitoring 9.e2
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
Supplementary Table 1. CD Variables, Univariate Logistic Regression Outcomes for Prediction of Active Disease (HBI>4)
Domain Variable n Cutoffs MLE SE Chi-square P value Modela Var#
Stool frequency Number of liquid/very soft stoolsper day
108 Continuous 1.63 0.35 22.34 <.0001 —
0/1–3/>3 3.49 0.73 23.11 <.0001 —
0/1–2/>2 3.55 0.66 28.58 <.0001 1 V1.10/>0 3.36 0.77 19.10 <.0001 —
How many stools did you haveyesterday during the day?
108 Continuous 0.85 0.19 20.09 <.0001 —
0–1/>1 2.00 0.65 9.53 .0020 —
Number of stools more thannormal per day
108 Continuous 0.87 0.36 5.88 .015 —
0/>0 0.84 0.51 2.69 .10 —
Total number of stools per day 108 Continuous 0.97 0.20 22.75 <.0001 —
�1/>1 1.75 0.77 5.08 .024 —
0–2/>2 2.58 0.55 22.19 <.0001 —
0–1/2–4/>4 2.64 0.58 20.79 <.0001 —
0–2/3–4/>4 2.16 0.40 28.49 <.0001 2 V1.2Stools at night How many stools did you have
last night?108 Continuous 0.88 0.26 11.59 .0007 —
0/>0 1.15 0.44 6.90 .0086 —
0/1/>1 0.99 0.30 11.01 .0009 —
0/1–2/>2 1.28 0.36 12.29 .0005 1, 2 V2Abdominal pain Abdominal pain (no–severe) 108 No/mild/moderate/severe 2.68 0.50 28.11 <.0001 —
No/yes 3.72 0.62 35.35 <.0001 1, 2 V3Abdominal pain (no pain–continuous) 108 No abdominal pain/with bowel actions/continuous 1.52 0.31 23.97 <.0001 —
Abdominal pain VAS 108 Continuous 0.51 0.10 24.99 <.0001 —
0–3/>3, <8/8–10 2.02 0.41 24.17 <.0001 —
<3/3–6/>6 1.94 0.39 25.03 <.0001 —
General well-being General well-being (very well– terrible) 108 Very well slightly below par/poor/very poor/terrible 1.31 0.30 18.93 <.0001 —
Well-being VAS 108 Continuous 0.47 0.11 18.44 <.0001 —
<4/�4, <7/�7 1.42 0.35 16.76 <.0001 —
�2/>2, <7/�7 1.80 0.40 20.09 <.0001 1, 2 V4Well-being (no impairment–unable
to work)108 No impairment/impaired but able to continue
activities/activities reduced/unable to work1.21 0.83 17.90 <.0001 —
Urgency Urgency of defecation(no urgency–incontinence)
108 No urgency/hurry/immediately/incontinence 1.30 0.32 16.99 <.0001 1, 2 V5No incontinence/incontinence 0.88 1.43 0.38 .54 —
Stool consistency Stool consistency 108 Normal or variably normal/semiformed/liquid 2.09 0.48 19.06 <.0001 1, 2 V6Do you take opiates or lomotil/imodium
for diarrhea?108 No/yes 1.22 0.52 5.47 .019 —
Antidiarrheals VAS 108 Continuous 0.00 0.09 0.00 1.00 —
Rectal bleeding Rectal bleeding VAS 108 Continuous 0.16 0.09 3.06 .080 —
�3/>3 1.22 0.52 5.47 .019 1, 2 V7What % of bowel movements contains
visible blood?108 None/<50%/�50%/blood alone 0.36 0.34 1.08 .30 —
Amount of blood in stool 108 None/trace/occasionally frank/usually frank 0.61 0.28 4.72 .030 —
Fever Did you have fever yesterday? 108 No/yes 14.53 913.40 0.00 .99 —
Anorexia Loss of appetite 108 No/yes 1.04 0.45 5.51 .019 1, 2 V8Nausea/vomiting Nausea and/or vomiting 108 No/yes 1.03 0.49 4.44 .035 1, 2 V9
FLA
5.4.0
DTD
�YJC
GH54544_proof�
11Decem
ber
2015�
12:31pm
�ce
CJ
9.e3Van
Deen
etal
Clinical
Gastroenterology
andHepatology
Vol.
-,No.
-
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
Disease activity Disease control VAS 108 Continuous 0.42 0.09 21.88 <.0001 —
�3, >3, <7/�7 1.60 0.32 24.41 <.0001 —
�2, >2, <7/�7 1.80 0.36 24.58 <.0001 1, 2 V10Laboratory studies C-reactive protein (mg/dL) 93 Continuous 0.27 0.14 3.80 .051 —
�0.8/>0.8 1.77 0.53 11.11 .0009 —
Calprotectin (mg/g) 63 Continuous 0.00 0.00 6.45 .011 —
<163/�163 1.22 0.63 3.71 .054 —
<50/�50 0.28 0.66 0.18 .67 —
White blood cell count (*103 cells/mL) 96 Continuous 0.15 0.08 3.20 .074 —
�9.95/>9.95 1.06 0.66 2.60 .11 —
Albumin (g/dL) 94 Continuous -1.55 0.61 6.47 .011 —
�3.7/<3.7 1.92 0.90 4.51 .034 —
Platelets (*103 cells/mL) 96 Continuous 0.00 0.00 1.43 .23 —
�143/<143 15.17 770.20 0.00 .98 —
Erythrocyte sedimentation rate (mm/hr) 85 Continuous 0.05 0.02 7.28 .0070 —
�22 (F) or �10 (M)/>22 (F) or >10 (M) 1.62 0.53 9.38 .0022 —
Hemoglobin (g/dL) 95 Continuous -0.10 0.14 0.59 .44 —
>11.6 (F) or >13.5 (M)/�11.6 (F) or �13.5 (M) 1.15 0.50 5.24 .022 —
Hematocrit 96 Continuous -3.34 5.39 0.38 .54 —
>0.349 (F) or >0.385 (M)/�0.349 (F) or �0.385 (M) 0.61 0.57 1.15 .28 —
F, female; M, male; MLE, maximum likelihood estimate, SE, standard error.aVariables selected for multivariate regression models (Supplementary Table 4).
FLA
5.4.0
DTD
�YJC
GH54544_proof�
11Decem
ber
2015�
12:31pm
�ce
CJ
-2015
mHealth
Indexfor
IBDMonitoring
9.e4
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
Supplementary Table 2. UC Variables, Univariate Logistic Regression Outcomes for Prediction of Active Disease (pMayo >2)
Domain Variable n Cutoffs MLE SE Chi-square P value Modela Var#
Stool frequency How many stools did you haveyesterday during the day?
109 Continuous 0.87 0.79 20.57 <.0001 —
<3/3–4/5–6/7–9/>9 1.61 0.33 23.11 <.0001 —
<4/4–6/>6 2.24 0.50 19.69 <.0001 —
<4/4–6/>6 2.32 0.49 22.35 <.0001 —
Stools more than normal 109 Continuous 0.99 0.33 9.23 .0024 —
0/1–2/3–4/>4 1.65 0.41 16.17 <.0001 —
0/1–2/>2 1.78 0.41 19.18 <.0001 —
Total number of stools/day 109 Continuous 0.78 0.16 23.03 <.0001 —
<3/3–6/>6 2.24 0.42 28.86 <.0001 —
0–4/>4 3.57 0.62 32.79 <.0001 1, 2 V1.10–2/3–4/>4 1.95 0.34 32.46 <.0001 3, 4 V1.2
Stools at night How many stools did youhave last night?
109 Continuous 0.95 0.26 12.77 .0004 —
0/>0 1.48 0.43 11.79 .0006 —
0/1–3/>3 1.49 0.38 15.01 .0001 1, 2, 3, 4 V2Abdominal pain Abdominal pain (no– severe) 109 No/mild/moderate/severe 0.78 0.25 9.60 .0019 —
No/yes 1.47 0.45 10.62 .0011 —
Abdominal pain (no– continuous) No abdominal pain 1.02 0.30 11.44 .0001 —
Abdominal pain VAS 109 Continuous 0.40 0.09 19.21 <.0001 —
<3/�3, �6/>6 1.75 0.37 21.84 <.0001 1, 2, 3, 4 V30/>0, �4/>4 1.40 0.33 18.20 <.0001 —
General well-being General well-being? (very well–terrible) 109 Very well/slightly below par/poor/very poor/terrible 1.27 0.34 14.12 .0002 —
General well-being (perfect–terrible) 109 Perfect/very good/good/average/poor/terrible 0.91 0.23 16.08 <.0001 1, 3 V4.1Well-being VAS 109 Continuous 0.43 0.11 15.30 <.0001 —
�3/>3 1.62 0.45 13.31 .0003 —
�3/>3, �6/>6 1.42 0.36 15.73 <.0001 2, 4 V4.2Well-being (no impairment–unable
to work)109 No impairment/impaired but able to continue
activities/activities reduced/unable to work1.14 0.36 15.93 <.0001 —
Urgency Urgency of defecation 109 No urgency/hurry/immediate/incontinence 2.01 0.40 24.82 <.0001 —
No urgency/hurry/immediate 2.03 0.40 25.13 <.0001 1, 2, 3, 4 V5No incontinence/incontinence 14.13 826.90 0.00 .99 —
Stool consistency Stool consistency 109 Normal or variably normal/semiformed/liquid 2.13 0.46 21.68 <.0001 1, 2, 3, 4 V6Do you take opiates or
lomotil/imodium for diarrhea?109 No/yes -0.03 0.59 0.00 .96 —
Antidiarrheals VAS 109 Continuous 0.05 0.09 0.26 .26 —
Rectal bleeding Rectal bleeding VAS 109 Continuous 0.59 0.11 27.82 <.0001 —
�3/>3 2.71 0.49 29.97 <.0001 —
<2/�2, <5/�5, <8/�8 1.66 0.31 28.15 <.0001 1, 2, 3, 4 V7What % of bowel movements
contains visible blood?109 None/<50%/�50%/blood alone 2.84 0.59 22.89 <.0001 —
None/<50%/�50% 2.95 0.60 24.26 <.0001 —
None/>0% 3.86 1.04 13.76 .0002 —
Amount of blood in stool 109 None/trace/occasionally frank/usually frank 1.83 0.35 27.22 <.0001 —
None/trace/more than a trace 1.95 0.37 27.53 <.0001 —
Anorexia Loss of appetite 109 No/yes 1.44 0.48 9.10 .0025 1, 2, 3, 4 V8Nausea/vomiting Nausea and/or vomiting 109 No/yes 0.72 0.49 2.12 .14 —
FLA
5.4.0
DTD
�YJC
GH54544_proof�
11Decem
ber
2015�
12:31pm
�ce
CJ
9.e5Van
Deen
etal
Clinical
Gastroenterology
andHepatology
Vol.
-,No.
-
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
Disease activity Disease control VAS 109 Continuous 0.46 0.10 21.67 <.0001 —
�3/>3, <7, �7 1.44 0.31 21.37 <.0001 —
�2/>2, �5/>5 1.82 0.36 25.86 <.0001 1, 2, 3, 4 V10�2/>2, <7/�7 1.68 0.36 21.76 <.0001 —
Labs C-reactive protein (mg/dL) 91 Continuous 1.03 0.42 5.87 .015 —
�0.8/>0.8 1.50 0.55 7.35 .0067 —
Calprotectin (mg/g) 70 Continuous 0.00 0.00 4.31 .038 —
<163/�163 1.29 0.56 5.22 .022 —
<50/�50 1.96 0.80 6.06 .014 —
White blood cell count (*103 cells/mL) 93 Continuous 0.20 0.08 5.91 .015 —
�9.95/>9.95 1.39 0.56 6.19 .013 —
Albumin (g/dL) 88 Continuous -1.52 0.66 5.39 .02 —
�3.7/<3.7 14.36 610.50 0.00 .98 —
Platelets (*103 cells/mL) 93 Continuous 0.00 0.00 2.54 .11 —
�143/<143 -12.66 805.50 0.00 .99 —
Erythrocyte sedimentation rate (mm/hr) 85 Continuous 0.04 0.01 8.38 .0038 —
�22 (F) or �10 (M)/>22 (F) or >10 (M) 1.04 0.47 4.81 .028 —
Hemoglobin (g/dL) 93 Continuous -0.31 0.15 4.59 .032 —
>11.6 (F) or >13.5 (M)/�11.6 (F) or �13.5 (M) 0.75 0.50 2.25 .13 —
Hematocrit 93 Continuous -9.46 5.91 2.66 .10 —
>0.349 (F) or >0.385 (M)/�0.349 (F) or �0.385 (M) 1.84 0.65 7.99 .0047 —
F, female; M, male; MLE, maximum likelihood estimate, SE, standard error.aVariables selected for multivariate regression models (Supplementary Table 4).
FLA
5.4.0
DTD
�YJC
GH54544_proof�
11Decem
ber
2015�
12:31pm
�ce
CJ
-2015
mHealth
Indexfor
IBDMonitoring
9.e6
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
Supplementary Table 3. Development and Evaluation of Composite Scores for CD
Composite score development AUC ROC curves P value
Dependent variable Model Variables in composite scoresa HBI CDAI HBI CDAI
HBI 1 V1.1; V3 0.981 0.912 0.903 0.7392 V1.2; V3; V10 0.956 0.898 0.750 0.702
CDAI 1 V1.1; V3; V4; V10 0.951 0.963 0.837 0.8302 V3; V4; V10 0.900 0.942 0.731 0.771
V1.1, number of liquid/very soft stools per day; V1.2, total number of stools per day; V3, abdominal pain; V4, well-being VAS; V10, disease control VAS.aSee Supplementary Table 1 for full details on each variable.
Supplementary Table 4. Development and Evaluation of Composite Scores for UC
Composite score development AUC ROC curves P value
Dependent variable Model Variables in composite scoresa pMayo SCCAI MTWI pMayo SCCAI MTWI
pMayo 1, 2 V1.1; V3; V7; V8; V10 0.960 0.865 0.849 0.769 0.769 0.7031, 2b V1.1; V3; V7; V10 0.957 0.879 0.883 0.797 0.803 0.762
3 V1.2; V3; V7; V8; V10 0.964 0.908 0.890 0.808 0.812 0.7483b V1.2; V3; V7; V10 0.960 0.915 0.913 0.820 0.832 0.7904 V1.2; V 4.2; V7; V10 0.956 0.920 0.909 0.809 0.838 0.787
SCCAI 1 V2; V4.1; V5; V7 0.872 0.974 0.896 0.711 0.907 0.8362, 4 V2; V3; V4.2; V5; V7; V10 0.904 0.971 0.883 0.765 0.911 0.810
3 V1.2; V2; V4.1; V5; V7 0.885 0.981 0.923 0.721 0.914 0.869MTWI 1 Unbalanced model NA NA NA NA NA NA
2 V1.1; V2; V6; V7; V8 0.928 0.879 0.937 0.801 0.806 0.8553 V1.2; V2; V4.1 0.880 0.907 0.984 0.712 0.797 0.9334 V1.2; V2; V4.2; V7; V8 0.906 0.894 0.950 0.777 0.810 0.866
MTWI, modified Truelove and Witts index; NA Q11; V1.1 and V1.2, number of stools per day; V2, number of stools at night; V3, abdominal pain VAS; V4.1, generalwell-being; V4.2, well-being VAS; V5, urgency of defecation; V6, stool consistency; V7, rectal bleeding VAS; V8, anorexia; V10, disease control VAS.aSee Supplementary Table 2 for full details on each variable.bIn these models loss of appetite was excluded as independent variable because of a clinically irrelevant negative value in the model.
FLA 5.4.0 DTD � YJCGH54544_proof � 11 December 2015 � 12:31 pm � ce CJ
9.e7 Van Deen et al Clinical Gastroenterology and Hepatology Vol. -, No. -
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
本文献由“学霸图书馆-文献云下载”收集自网络,仅供学习交流使用。
学霸图书馆(www.xuebalib.com)是一个“整合众多图书馆数据库资源,
提供一站式文献检索和下载服务”的24 小时在线不限IP
图书馆。
图书馆致力于便利、促进学习与科研,提供最强文献下载服务。
图书馆导航:
图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具