-
Notifications
You must be signed in to change notification settings - Fork 0
/
Analyses.do
558 lines (427 loc) · 33.7 KB
/
Analyses.do
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
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
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
232
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
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
464
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
/** ANALYSES FOR THE DENSITY ARTICLE IN THE TO-BE TRIAL **/
/* ---------- Setup ---------- */
clear all
cd "M:\Mammografi\Prosjekter\02 Tomosyntese\C Hordaland\Prosjekter\TOBE-1 - artikkel 4\"
use Analysefil_artikkel4.dta, clear
// Some labels
label define ScreeningType 1 "DM" 2 "DBT"
label value ScreeningType ScreeningType
label define VBD_gr 0 "First quintile" 1 "Second quintile" 2 "Third quintile" 3 "Fourth quintile" 4 "Fifth quintile"
label value VBD_gr VBD_gr
label define VDG_5ed 1 "VDG 1" 2 "VDG 2" 3 "VDG 3" 4 "VDG 4"
label value VDG_5ed VDG_5ed
// Global graph scheme for prettier graphs
set scheme s1mono
/* ---------- Table 1 ---------- */
// Descriptive statistics with median and IQR for continuous variables and percentages for categorical variables
table ScreeningType, c(p50 alder p25 alder p75 alder) // age
tab ScreeningType pre_sub, row // screening history
table ScreeningType, c(p50 BMI p25 BMI p75 BMI) // BMI
table ScreeningType, c(p50 BV_cm3 p25 BV_cm3 p75 BV_cm3) // Breast volume
table ScreeningType, c(p50 VBD p25 VBD p75 VBD) // Volumetric breast density
table ScreeningType, c(p50 FGV_cm3 p25 FGV_cm3 p75 FGV_cm3) // Fibroglandular volume
table ScreeningType, c(p50 H_mm p25 H_mm p75 H_mm) // Compressed breast thickness
tab ScreeningType VDG_5ed, row // Volpara Density Grades
table VDG_5ed ScreeningType, c(min VBD max VBD)
tab ScreeningType VBD_gr, row // Volumetric breast density quintiles
table VBD_gr ScreeningType, c(min VBD max VBD)
/* ---------- Figure 2 ---------- */
// Bar graphs with 95 % confidence interval around the mean. Texts are overlaid through a combination of relabeling and scatter plotting strings directly onto the graph.
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p=0.001""' 2 `""VDG 2" "n=6216 n=6280" "p=0.002""' 3 `""VDG 3" "n=3152 n=3655" "p=0.88""' 4 `""VDG 4" " n=962 n=1136" "p=0.26""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanrecall = recall (sd) sdrecall = recall (count) n = recall, by(ScreeningType VDG_5ed)
gen prosentrecall = string(100 * meanrecall, "%8.1f") + "%"
gen hirecall = meanrecall + invttail(n-1,0.025)*(sdrecall / sqrt(n))
gen lorecall = meanrecall - invttail(n-1,0.025)*(sdrecall / sqrt(n))
graph twoway (bar meanrecall ScreeningType, fcolor(red%40)) (rcap hirecall lorecall ScreeningType, lcolor(blue%40)) (scatter meanrecall ScreeningType, ms(none) mlabel(prosentrecall) mlabpos(6)), ///
by(VDG_5ed, title("A: Recall", ///
span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white) margin(medium)) ///
yscale(range(0 0.05)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("Recall (%)") ///
ylabel(0(0.01)0.05 0 "0" 0.01 "1" 0.02 "2" 0.03 "3" 0.04 "4" 0.05 "5", grid angle(h)) plotregion(color(white)) graphregion(color(white))
graph save figurrecall.gph, replace
restore
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p<0.001""' 2 `""VDG 2" "n=6216 n=6280" "p<0.001""' 3 `""VDG 3" "n=3152 n=3655" "p=0.95""' 4 `""VDG 4" " n=962 n=1136" "p=0.21""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanFP = FP (sd) sdFP = FP (count) n = FP, by(ScreeningType VDG_5ed)
gen prosentFP = string(100 * meanFP, "%8.1f") + "%"
gen hiFP = meanFP + invttail(n-1,0.025)*(sdFP / sqrt(n))
gen loFP = meanFP - invttail(n-1,0.025)*(sdFP / sqrt(n))
graph twoway (bar meanFP ScreeningType, fcolor(red%40)) (rcap hiFP loFP ScreeningType, lcolor(blue%40)) (scatter meanFP ScreeningType, ms(none) mlabel(prosentFP) mlabpos(6)), ///
by(VDG_5ed, title("B: False positives" ///
, span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white)) ///
yscale(range(0 0.05)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("False positives (%)") ///
ylabel(0(0.01)0.05 0 "0" 0.01 "1" 0.02 "2" 0.03 "3" 0.04 "4" 0.05 "5", grid angle(h)) plotregion(color(white)) graphregion(color(white))
graph save figurFP.gph, replace
restore
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p=0.44""' 2 `""VDG 2" "n=6216 n=6280" "p=0.10""' 3 `""VDG 3" "n=3152 n=3655" "p=0.10""' 4 `""VDG 4" " n=962 n=1136" "p=0.72""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanbiopsi = biopsi (sd) sdbiopsi = biopsi (count) n = biopsi, by(ScreeningType VDG_5ed)
gen prosentbiopsi = string(100 * meanbiopsi, "%8.1f") + "%"
gen hibiopsi = meanbiopsi + invttail(n-1,0.025)*(sdbiopsi / sqrt(n))
gen lobiopsi = meanbiopsi - invttail(n-1,0.025)*(sdbiopsi / sqrt(n))
graph twoway (bar meanbiopsi ScreeningType, fcolor(red%40)) (rcap hibiopsi lobiopsi ScreeningType, lcolor(blue%40)) (scatter meanbiopsi ScreeningType, ms(none) mlabel(prosentbiopsi) mlabpos(6)), ///
by(VDG_5ed, title("C: Biopsy" ///
, span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white)) ///
yscale(range(0 0.03)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("Biopsy (%)") ///
ylabel(0(0.005)0.03 0 "0" 0.005 "0.5" 0.01 "1" 0.015 "1.5" 0.02 "2" 0.025 "2.5" 0.03 "3", grid angle(h)) plotregion(color(white)) graphregion(color(white))
graph save figurbiopsi.gph, replace
restore
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p=0.96""' 2 `""VDG 2" "n=6216 n=6280" "p=0.31""' 3 `""VDG 3" "n=3152 n=3655" "p=0.82""' 4 `""VDG 4" " n=962 n=1136" "p=0.98""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanSDC = SDC (sd) sdSDC = SDC (count) n = SDC, by(ScreeningType VDG_5ed)
gen prosentSDC = string(100 * meanSDC, "%8.2f") + "%"
gen hiSDC = meanSDC + invttail(n-1,0.025)*(sdSDC / sqrt(n))
gen loSDC = meanSDC - invttail(n-1,0.025)*(sdSDC / sqrt(n))
graph twoway (bar meanSDC ScreeningType, fcolor(red%40)) (rcap hiSDC loSDC ScreeningType, lcolor(blue%40)) (scatter meanSDC ScreeningType, ms(none) mlabel(prosentSDC) mlabpos(6)), ///
by(VDG_5ed, title("D: Screen-detected cancers" ///
, span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white)) ///
yscale(range(0 0.015)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("Screen-detected cancers (%)") ///
ylabel(0(0.002)0.014 0 "0" 0.002 "0.2" 0.004 "0.4" 0.006 "0.6" 0.008 "0.8" 0.01 "1" 0.012 "1.2" 0.014 "1.4", grid angle(h)) plotregion(color(white)) graphregion(color(white))
graph save figurSDC.gph, replace
restore
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p=0.15""' 2 `""VDG 2" "n=6216 n=6280" "p=0.01""' 3 `""VDG 3" "n=3152 n=3655" "p=0.86""' 4 `""VDG 4" " n=962 n=1136" "p=0.59""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanrecall = recall (mean) meanSDC = SDC (count) n = recall if recall == 1, by(ScreeningType VDG_5ed)
gen meanPPV1 = meanSDC/meanrecall
gen prosentPPV1 = string(100 * meanPPV1, "%8.1f") + "%"
gen hiPPV1 = meanPPV1 + 1.96*(sqrt((meanPPV1*(1-meanPPV1))/n))
gen loPPV1 = meanPPV1 - 1.96*(sqrt((meanPPV1*(1-meanPPV1))/n))
graph twoway (bar meanPPV1 ScreeningType, fcolor(red%40)) (rcap hiPPV1 loPPV1 ScreeningType, lcolor(blue%40)) (scatter meanPPV1 ScreeningType, ms(none) mlabel(prosentPPV1) mlabpos(6)), ///
by(VDG_5ed, title("E: Positive Predictive Value-1", ///
span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white)) ///
yscale(range(0 0.5)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("Positive Predictive Value-1 (%)") ///
ylabel(0(0.1)0.5 0 "0" 0.1 "10" 0.2 "20" 0.3 "30" 0.4 "40" 0.5 "50", grid angle(h)) plotregion(color(white)) graphregion(color(white))
graph save figurPPV1.gph, replace
restore
preserve
replace ScreeningType = 0 if ScreeningType == 2
label define VDG_5ed2 1 `""VDG 1" "n=3929 n=3212" "p=0.62""' 2 `""VDG 2" "n=6216 n=6280" "p=0.01""' 3 `""VDG 3" "n=3152 n=3655" "p=0.38""' 4 `""VDG 4" " n=962 n=1136" "p=0.80""'
label value VDG_5ed VDG_5ed2
collapse (mean) meanbiopsi = recall (mean) meanSDC = SDC (count) n = biopsi if biopsi == 1, by(ScreeningType VDG_5ed)
gen meanPPV2 = meanSDC/meanbiopsi
gen prosentPPV2 = string(100 * meanPPV2, "%8.1f") + "%"
gen hiPPV2 = meanPPV2 + 1.96*(sqrt((meanPPV2*(1-meanPPV2))/n))
gen loPPV2 = meanPPV2 - 1.96*(sqrt((meanPPV2*(1-meanPPV2))/n))
graph twoway (bar meanPPV2 ScreeningType, fcolor(red%40)) (rcap hiPPV2 loPPV2 ScreeningType, lcolor(blue%40)) (scatter meanPPV2 ScreeningType, ms(none) mlabel(prosentPPV2) mlabpos(6)), ///
by(VDG_5ed, title("F: Positive Predictive Value-3" ///
, span size(5)) subtitle("Volpara Density Grades (VDG)") rows(1) legend(off) note("")) subtitle(,fcolor(white) bcolor(white)) ///
yscale(range(0 0.75)) xlabel(0 "DBT" 1 "DM", noticks) xtitle("Screening technique") ytitle("Positive Predictive Value-3 (%)") ///
ylabel(0(0.1)0.75 0 "0" 0.1 "10" 0.2 "20" 0.3 "30" 0.4 "40" 0.5 "50" 0.6 "60" 0.7 "70", grid angle(h)) ///
plotregion(color(white)) graphregion(color(white))
graph save figurPPV2.gph, replace
restore
gr combine figurrecall.gph figurFP.gph figurbiopsi.gph figurSDC.gph figurPPV1.gph figurPPV2.gph, iscale(0.33) col(2) caption("DBT: Digital Breast Tomosynthesis, DM: Digital Mammography", size(half_tiny))
graph export figur2.png, replace height(4000) width(3200)
graph export figur2_tif.tif, replace height(4000) width(3200)
erase figurFP.gph
erase figurrecall.gph
erase figurbiopsi.gph
erase figurSDC.gph
erase figurPPV1.gph
erase figurPPV2.gph
/* ---------- Table 2 and 3 ---------- */
// Histopathological tumour characteristics
// Tumour size variable replaced with a more precise one from a radiological review
replace SLdm_solitærettf = "" if SLdm_solitærettf == "kalk"
destring SLdm_solitærettf, replace
replace Tumorstr = SLdm_solitærettf if SLdm_solitærettf != .
// Crude proportions
tab Gruppe if Gruppe==3 & ScreeningType == 1 & VDG_5ed != .
tab Gruppe if Gruppe==3 & ScreeningType == 2 & VDG_5ed != .
table VDG_5ed if Gruppe==3 & ScreeningType == 1 & Tumorstr<999, c(median Tumorstr iqr Tumorstr p25 Tumorstr p75 Tumorstr)
table VDG_5ed if Gruppe==3 & ScreeningType == 2 & Tumorstr<999, c(median Tumorstr iqr Tumorstr p25 Tumorstr p75 Tumorstr)
tab Grad VDG_5ed if Gruppe==3 & ScreeningType == 1, col
tab Grad VDG_5ed if Gruppe==3 & ScreeningType == 2, col
tab pN VDG_5ed if Gruppe==3 & ScreeningType == 1, col
tab pN VDG_5ed if Gruppe==3 & ScreeningType == 2, col
tab subtype VDG_5ed if Gruppe==3 & ScreeningType == 1, col mi
tab subtype VDG_5ed if Gruppe==3 & ScreeningType == 2, col mi
// Confidence intervals
by VDG_5ed : ci mean Tumorstr if ScreeningType == 1 & Tumorstr<999 & Gruppe==3, total
by VDG_5ed : ci mean Tumorstr if ScreeningType == 2 & Tumorstr<999 & Gruppe==3, total
proportion Grad if ScreeningType == 1 & Grad != 9 & VDG_5ed != ., citype(exact) percent
proportion Grad if ScreeningType == 2 & Grad != 9 & VDG_5ed != ., citype(exact) percent
proportion Grad if ScreeningType == 1 & Grad != 9 & VDG_5ed != ., citype(exact) over(VDG_5ed) percent
proportion Grad if ScreeningType == 2 & Grad != 9 & VDG_5ed != ., citype(exact) over(VDG_5ed) percent
proportion pN if ScreeningType == 1 & pN != 999 & VDG_5ed != . & Gruppe==3, citype(exact) percent
proportion pN if ScreeningType == 2 & pN != 999 & VDG_5ed != . & Gruppe==3, citype(exact) percent
proportion pN if ScreeningType == 1 & pN != 999 & VDG_5ed != . & Gruppe==3, citype(exact) over(VDG_5ed) percent
proportion pN if ScreeningType == 2 & pN != 999 & VDG_5ed != . & Gruppe==3, citype(exact) over(VDG_5ed) percent
proportion subtype if ScreeningType == 1 & subtype != . & VDG_5ed != . & Gruppe==3, citype(exact) percent
proportion subtype if ScreeningType == 2 & subtype != . & VDG_5ed != . & Gruppe==3, citype(exact) percent
proportion subtype if ScreeningType == 1 & subtype != . & VDG_5ed != . & Gruppe==3, citype(exact) over(VDG_5ed) percent
proportion subtype if ScreeningType == 2 & subtype != . & VDG_5ed != . & Gruppe==3, citype(exact) over(VDG_5ed) percent
/* ---------- Table 4a-c ---------- */
// relative risks produced through log-binomial regression models with exponentiated coefficients for recall, false positiv and screen-detected cancers by density, adjusting for a continuous measure of breast volume.
// All models run unadjusted before final adjusted models.
glm recall i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm recall i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm FP i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
glm SDC i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(log) eform
/* ---------- Supplemental material ---------- */
// Supplemental table S1-S3 - Sensitivity analysis
glm SDC i.VBD_gr if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VBD_gr if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VBD_gr if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VBD_gr if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VBD_gr if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VBD_gr if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VBD_gr i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VBD_gr i.alder_gr i.pre_sub BMI if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VBD_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VBD_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VBD_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VBD_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VBD_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VBD_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VBD_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VDG_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm SDC i.VDG_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm SDC i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VDG_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm recall i.VDG_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm recall i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VDG_tokat if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1, family(binomial) link(log) eform
glm FP i.VDG_tokat if ScreeningType == 2, family(binomial) link(log) eform
glm FP i.VDG_tokat i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2, family(binomial) link(log) eform
/* ---------- Supplemental material, Table S4 ---------- */
// The numbers behind the bar graphs
// Code for Z tests for proportions looped over categories and performance measures provide CIs
foreach var of varlist konsensus FP recall biopsi SDC {
prtest `var' if VDG_5ed == 1, by(ScreeningType)
}
foreach var of varlist konsensus FP recall biopsi SDC {
prtest `var' if VDG_5ed == 2, by(ScreeningType)
}
foreach var of varlist konsensus FP recall biopsi SDC {
prtest `var' if VDG_5ed == 3, by(ScreeningType)
}
foreach var of varlist konsensus FP recall biopsi SDC {
prtest `var' if VDG_5ed == 4, by(ScreeningType)
}
// Crude threeway tables
table recall VDG_5ed ScreeningType
table biopsi VDG_5ed ScreeningType
table SDC VDG_5ed ScreeningType
// Immediate code for Z tests for PPV
prtesti 81 18 106 15, c // VDG 1 - PPV-1
prtesti 48 18 46 15, c // VDG 1 - PPV-3
prtesti 200 48 267 39, c // VDG 2 - PPV-1
prtesti 103 48 129 39, c // VDG 2 - PPV-3
prtesti 129 23 147 25, c // VDG 3 - PPV-1
prtesti 78 23 69 25, c // VDG 3 - PPV-3
prtesti 30 6 46 7, c // VDG 4 - PPV-1
prtesti 19 6 25 7, c // VDG 4 - PPV-3
/* ---------- Supplemental material, Table S5 ---------- */
// Correlation matrix
pwcorr BMI BV_cm3 VBD FGV_cm3, sig
/* ---------- Supplemental material, Table S6 ---------- */
// Log-binomial regression models for DM and DBT combined using interaction terms
glm recall i.VDG_5ed, family(binomial) link(log) eform allbase
glm recall i.ScreeningType, family(binomial) link(log) eform allbase
glm recall i.alder_gr, family(binomial) link(log) eform allbase
glm recall i.pre_sub, family(binomial) link(log) eform allbase
glm recall BV_cm3, family(binomial) link(log) eform allbase
glm recall i.VDG_5ed##i.ScreeningType i.alder_gr i.pre_sub BV_cm3, family(binomial) link(log) eform allbase
glm FP i.VDG_5ed, family(binomial) link(log) eform allbase
glm FP i.ScreeningType, family(binomial) link(log) eform allbase
glm FP i.alder_gr, family(binomial) link(log) eform allbase
glm FP i.pre_sub, family(binomial) link(log) eform allbase
glm FP BV_cm3, family(binomial) link(log) eform allbase
glm FP i.VDG_5ed##i.ScreeningType i.alder_gr i.pre_sub BV_cm3, family(binomial) link(log) eform allbase
glm SDC i.VDG_5ed, family(binomial) link(log) eform allbase
glm SDC i.ScreeningType, family(binomial) link(log) eform allbase
glm SDC i.alder_gr, family(binomial) link(log) eform allbase
glm SDC i.pre_sub, family(binomial) link(log) eform allbase
glm SDC BV_cm3, family(binomial) link(log) eform allbase
glm SDC i.VDG_5ed##i.ScreeningType i.alder_gr i.pre_sub BV_cm3, family(binomial) link(log) eform allbase
/* ---------- Supplemental material, Figure S1 ---------- */
// Histogram
twoway (histogram VBD if ScreeningType == 1, lcolor(blue%0) color(blue%30) start(1.5) width(1) freq) (histogram VBD if ScreeningType == 2, freq lcolor(red%0) color(red%30) width(1) start(1.5)), legend(order(1 "DM" 2 "DBT"))
graph export Figur4.png, width(800) height(600) replace
graph export Figur4_tif.tif, width(800) height(600) replace
/* _____________________________________________________________________________________________________________ */
/* ---------- UNUSED ADDITIONAL ANALYSES ---------- */
// Alternative logistic regression models
glm recall i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm recall i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm FP i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.VDG_5ed if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.alder_gr if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.pre_sub if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 1 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.VDG_5ed if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.alder_gr if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.pre_sub if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
glm SDC i.VDG_5ed i.alder_gr i.pre_sub BV_cm3 if ScreeningType == 2 & VDG_5ed != . & alder_gr != . & pre_sub != . & BV_cm3 != ., family(binomial) link(logit) eform
// Graph using splines for SDC over a continuous VBD variable
mkspline VBDcubDM = VBD if ScreeningType == 1, cubic nknots(4)
glm SDC VBDcubDM* i.alder_gr i.pre_sub BV_cm3, family(binomial) link(logit) eform
estat ic
predictnl xb1 = _b[_cons] + _b[VBDcubDM1]*VBDcubDM1 + _b[VBDcubDM2]*VBDcubDM2 + _b[VBDcubDM3]*VBDcubDM3, se(se1)
gen lb_xb1 = xb1 - invnormal(0.975)*se1
gen ub_xb1 = xb1 + invnormal(0.975)*se1
gen lb1 = exp(lb_xb1)/(1+exp(lb_xb1))
gen ub1 = exp(ub_xb1)/(1+exp(ub_xb1))
gen or1 = exp(xb1)
gen p1 = or1 / (1 + or1)
mkspline VBDcubDBT = VBD if ScreeningType == 2, cubic nknots(4)
glm SDC VBDcubDBT* i.alder_gr i.pre_sub BV_cm3, family(binomial) link(logit) eform
estat ic
predictnl xb2 = _b[_cons] + _b[VBDcubDBT1]*VBDcubDBT1 + _b[VBDcubDBT2]*VBDcubDBT2 + _b[VBDcubDBT3]*VBDcubDBT3, se(se2)
gen lb_xb2 = xb2 - invnormal(0.975)*se2
gen ub_xb2 = xb2 + invnormal(0.975)*se2
gen lb2 = exp(lb_xb2)/(1+exp(lb_xb2))
gen ub2 = exp(ub_xb2)/(1+exp(ub_xb2))
gen or2 = exp(xb2)
gen p2 = or2 / (1 + or2)
bysort VBD: gen merk = 1 if _n == 1
tw (rarea lb1 ub1 VBD if VBD <= 20, color(blue%30) sort) (rarea lb2 ub2 VBD if VBD <= 20, color(red%30) sort) (line p1 VBD if VBD <= 20, sort) (line p2 VBD if VBD <= 20, sort), xtitle("Volumetric Breast Density") ytitle("SDC") legend(order(1 "DM" 2 "DBT")) xlabel(2(2)20) ylabel(.0 "0 %" .004 "0.4 %" .008 "0.8 %" .012 "1.2 %" .016 "1.6 %", angle(h))
graph export SplinesSDC.png, width(800) height(600) replace
drop VBDcubDM1-p2
// Graph using splines for recall over a continuous VBD variable
mkspline VBDcubDM = VBD if ScreeningType == 1, cubic nknots(4)
glm recall VBDcubDM* i.alder_gr i.pre_sub BV_cm3, family(binomial) link(logit) eform
estat ic
predictnl xb1 = _b[_cons] + _b[VBDcubDM1]*VBDcubDM1 + _b[VBDcubDM2]*VBDcubDM2 + _b[VBDcubDM3]*VBDcubDM3, se(se1)
gen lb_xb1 = xb1 - invnormal(0.975)*se1
gen ub_xb1 = xb1 + invnormal(0.975)*se1
gen lb1 = exp(lb_xb1)/(1+exp(lb_xb1))
gen ub1 = exp(ub_xb1)/(1+exp(ub_xb1))
gen or1 = exp(xb1)
gen p1 = or1 / (1 + or1)
mkspline VBDcubDBT = VBD if ScreeningType == 2, cubic nknots(4)
glm recall VBDcubDBT* i.alder_gr i.pre_sub BV_cm3, family(binomial) link(logit) eform
estat ic
predictnl xb2 = _b[_cons] + _b[VBDcubDBT1]*VBDcubDBT1 + _b[VBDcubDBT2]*VBDcubDBT2 + _b[VBDcubDBT3]*VBDcubDBT3, se(se2)
gen lb_xb2 = xb2 - invnormal(0.975)*se2
gen ub_xb2 = xb2 + invnormal(0.975)*se2
gen lb2 = exp(lb_xb2)/(1+exp(lb_xb2))
gen ub2 = exp(ub_xb2)/(1+exp(ub_xb2))
gen or2 = exp(xb2)
gen p2 = or2 / (1 + or2)
tw (rarea lb1 ub1 VBD if VBD <= 20, color(blue%30) sort) (rarea lb2 ub2 VBD if VBD <= 20, color(red%30) sort) (line p1 VBD if VBD <= 20, sort) (line p2 VBD if VBD <= 20, sort), xtitle("Volumetric Breast Density") ytitle("Recall") legend(order(1 "DM" 2 "DBT")) xlabel(2(2)20) ylabel(.0 "0 %" .02 "2 %" .04 "4 %" .06 "6 %" .08 "8 %" .1 "10 %", angle(h))
graph export SplinesRecall.png, width(800) height(600) replace
drop VBDcubDM1-p2
// Crude spline graph, SDC over VBD
mkspline VBDcubDM = VBD if ScreeningType == 1, cubic nknots(4)
glm SDC VBDcubDM*, family(binomial) link(logit) eform
estat ic
predictnl xb1 = _b[_cons] + _b[VBDcubDM1]*VBDcubDM1 + _b[VBDcubDM2]*VBDcubDM2 + _b[VBDcubDM3]*VBDcubDM3, se(se1)
gen lb_xb1 = xb1 - invnormal(0.975)*se1
gen ub_xb1 = xb1 + invnormal(0.975)*se1
gen lb1 = exp(lb_xb1)/(1+exp(lb_xb1))
gen ub1 = exp(ub_xb1)/(1+exp(ub_xb1))
gen or1 = exp(xb1)
gen p1 = or1 / (1 + or1)
mkspline VBDcubDBT = VBD if ScreeningType == 2, cubic nknots(4)
glm SDC VBDcubDBT*, family(binomial) link(logit) eform
estat ic
predictnl xb2 = _b[_cons] + _b[VBDcubDBT1]*VBDcubDBT1 + _b[VBDcubDBT2]*VBDcubDBT2 + _b[VBDcubDBT3]*VBDcubDBT3, se(se2)
gen lb_xb2 = xb2 - invnormal(0.975)*se2
gen ub_xb2 = xb2 + invnormal(0.975)*se2
gen lb2 = exp(lb_xb2)/(1+exp(lb_xb2))
gen ub2 = exp(ub_xb2)/(1+exp(ub_xb2))
gen or2 = exp(xb2)
gen p2 = or2 / (1 + or2)
bysort VBD: gen merk = 1 if _n == 1
tw (rarea lb1 ub1 VBD if VBD <= 20, color(blue%30) sort) (rarea lb2 ub2 VBD if VBD <= 20, color(red%30) sort) (line p1 VBD if VBD <= 20, sort) (line p2 VBD if VBD <= 20, sort), xtitle("Volumetric Breast Density") ytitle("SDC") legend(order(1 "DM" 2 "DBT")) xlabel(2(2)20) ylabel(.0 "0 %" .004 "0.4 %" .008 "0.8 %" .012 "1.2 %" .016 "1.6 %", angle(h))
graph export SplinesSDC_crude.png, width(800) height(600) replace
drop VBDcubDM1-p2
// Crude spline graph, recall over VBD
mkspline VBDcubDM = VBD if ScreeningType == 1, cubic nknots(4)
glm recall VBDcubDM*, family(binomial) link(logit) eform
estat ic
predictnl xb1 = _b[_cons] + _b[VBDcubDM1]*VBDcubDM1 + _b[VBDcubDM2]*VBDcubDM2 + _b[VBDcubDM3]*VBDcubDM3, se(se1)
gen lb_xb1 = xb1 - invnormal(0.975)*se1
gen ub_xb1 = xb1 + invnormal(0.975)*se1
gen lb1 = exp(lb_xb1)/(1+exp(lb_xb1))
gen ub1 = exp(ub_xb1)/(1+exp(ub_xb1))
gen or1 = exp(xb1)
gen p1 = or1 / (1 + or1)
mkspline VBDcubDBT = VBD if ScreeningType == 2, cubic nknots(4)
glm recall VBDcubDBT*, family(binomial) link(logit) eform
estat ic
predictnl xb2 = _b[_cons] + _b[VBDcubDBT1]*VBDcubDBT1 + _b[VBDcubDBT2]*VBDcubDBT2 + _b[VBDcubDBT3]*VBDcubDBT3, se(se2)
gen lb_xb2 = xb2 - invnormal(0.975)*se2
gen ub_xb2 = xb2 + invnormal(0.975)*se2
gen lb2 = exp(lb_xb2)/(1+exp(lb_xb2))
gen ub2 = exp(ub_xb2)/(1+exp(ub_xb2))
gen or2 = exp(xb2)
gen p2 = or2 / (1 + or2)
tw (rarea lb1 ub1 VBD if VBD <= 20, color(blue%30) sort) (rarea lb2 ub2 VBD if VBD <= 20, color(red%30) sort) (line p1 VBD if VBD <= 20, sort) (line p2 VBD if VBD <= 20, sort), xtitle("Volumetric Breast Density") ytitle("Recall") legend(order(1 "DM" 2 "DBT")) xlabel(2(2)20) ylabel(.0 "0 %" .02 "2 %" .04 "4 %" .06 "6 %" .08 "8 %" .1 "10 %", angle(h))
graph export SplinesRecall_crude.png, width(800) height(600) replace
drop VBDcubDM1-p2