generated from opensafely/research-template
/
model.do
399 lines (296 loc) · 19.5 KB
/
model.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
/*======================================================================================
DO FILE NAME: model.do
PROJECT: Germ Defence
AUTHOR: S Walter
DATE: Sept 2021
DESCRIPTION OF FILE: Merge data for each outcome into one
Run intervention-control comparisons - as per intention-to-treat analysis
Run interrupted time series models as secondary analysis
-Negative binomial GLM
-Negative binomial random intercepts
DATASETS USED: data in memory ($output/measure/smeasure_*.csv)
DATASETS CREATED: output/practice_weekly.dta
output/measures/measure_*_weekly.dta x8
output/process_eval.dta
OTHER OUTPUT: PerProtocol.xlsx, DiD_weekly.xlsx
CITS_weekly.xlsx, CITS_RandInt.xlsx
*_plots.pdf x8
=======================================================================================*/
global gd "`c(pwd)'"
*convert CSV files to .dta files
import delimited using "$gd/output/measures/measure_RTI_weekly.csv", clear
drop value
save "$gd/output/measures/measure_RTI_weekly.dta", replace
import delimited using "$gd/output/measures/measure_aRTI_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_aRTI_weekly.dta", replace
import delimited using "$gd/output/measures/measure_gastro_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_gastro_weekly.dta", replace
import delimited using "$gd/output/measures/measure_coviddiag_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_coviddiag_weekly.dta", replace
import delimited using "$gd/output/measures/measure_covidsympsens_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_covidsympsens_weekly.dta", replace
import delimited using "$gd/output/measures/measure_covidsympspec_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_covidsympspec_weekly.dta", replace
import delimited using "$gd/output/measures/measure_antibio_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_antibio_weekly.dta", replace
import delimited using "$gd/output/measures/measure_adm_weekly.csv", clear
drop population value
save "$gd/output/measures/measure_adm_weekly.dta", replace
*get weekly level counts for any type of GP consultation - treat as base dataset
use "$gd/output/measures/measure_RTI_weekly.dta", clear
*+merge on tables for other outcomes to create single unified dataset
merge 1:1 practice_id date using "$gd/output/measures/measure_aRTI_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_gastro_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_coviddiag_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_covidsympsens_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_covidsympspec_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_antibio_weekly.dta"
drop _merge
merge 1:1 practice_id date using "$gd/output/measures/measure_adm_weekly.dta"
drop _merge
*fill in missing counts
replace rti_events=0 if rti_events==.|rti_events<0
replace arti_events=0 if arti_events==.|arti_events<0
replace gastro_events=0 if gastro_events==.|gastro_events<0
replace coviddiag_events=0 if coviddiag_events==.|coviddiag_events<0
replace covidsympsens_events=0 if covidsympsens_events==.|covidsympsens_events<0
replace covidsympspec_events=0 if covidsympspec_events==.|covidsympspec_events<0
replace antibio_events=0 if antibio_events==.|antibio_events<0
replace adm_events=0 if adm_events==.|adm_events<0
*define pre and post intervention periods
generate date2 = date(date, "YMD")
format %td date2
gen period = (date2>=d(10nov2020))
*consecutively number weeks
sort date2 practice_id
gen year = year(date2)
gen week = week(date2)
gen week_date = week - 27
replace week_date = week + 25 if year==2021
*save practice x week level data
save "$gd/output/practice_weekly.dta", replace
export excel using "$gd/output/practice_weekly.xlsx", firstrow(variables) replace
*** I. Intention-to-treat analysis
*collapse post-intervention data to one row per practice - sum event counts for each outcome
drop if period!=1
collapse (sum) rti_events arti_events gastro_events coviddiag_events covidsympsens_events covidsympspec_events antibio_events adm_events population, by(practice_id)
*add practice covariates
merge 1:1 practice_id using "$gd/output/practice_variables.dta"
drop _merge
*save data for use in process evaluation
save "$gd/output/process_eval.dta", replace
*Compare event rates between intervention vs. control during post-intervention period: 10/11/20 - 15/03/21
glm rti_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("RTI") replace
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm arti_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("aRTI") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm gastro_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("gastro") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm coviddiag_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("CovidDiag") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm covidsympsens_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("CovidSympSens") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm covidsympspec_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("CovidSympSpec") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm antibio_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("Antibio") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
glm adm_events i.intervention, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/PerProtocolv2.xlsx", sheet("adm") modify
putexcel A1=matrix(r(table)), names
lincom _cons + 1.intervention
*** II. Conrolled interrupted time series analysis
** A. Weekly level analysis
*get population values by intervention group
use "$gd/output/practice_variables.dta", replace
collapse (sum) list_size, by(intervention)
save "$gd/output/pop_by_intervention.dta", replace
*get practice x week level data, then collapse to one row per week, summing event counts for each outcome
use "$gd/output/practice_weekly.dta", replace
*add intervention group indicator
merge m:1 practice_id using "$gd/output/practice_variables.dta"
drop _merge
drop if intervention==.
collapse (sum) rti_events arti_events gastro_events coviddiag_events covidsympsens_events covidsympspec_events antibio_events adm_events population, by(date2 week_date intervention)
gen time=week_date - 17.5
gen period = (date2>=d(10nov2020))
*add population
merge m:1 intervention using "$gd/output/pop_by_intervention.dta"
drop _merge
*create interaction variables for CITS model [as described in Linden (2015) The Stata Journal, 15(2), pp. 480–500]
gen time_period = 0
replace time_period = time if period==1
gen time_intervention = 0
replace time_intervention = time if intervention==1
gen intervention_period = 0
replace intervention_period = 1 if intervention==1 & period==1
gen time_int_period = 0
replace time_int_period = time if intervention_period==1
* i. Difference in differences model
glm rti_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("RTI") replace
putexcel A1=matrix(r(table)), names
glm arti_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("aRTI") modify
putexcel A1=matrix(r(table)), names
glm gastro_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("gastro") modify
putexcel A1=matrix(r(table)), names
glm coviddiag_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("CovidDiag") modify
putexcel A1=matrix(r(table)), names
glm covidsympsens_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("CovidSympSens") modify
putexcel A1=matrix(r(table)), names
glm covidsympspec_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("CovidSympSpec") modify
putexcel A1=matrix(r(table)), names
glm antibio_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("antibio") modify
putexcel A1=matrix(r(table)), names
glm adm_events intervention period intervention_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/DiD_weekly.xlsx", sheet("adm") modify
putexcel A1=matrix(r(table)), names
* ii. Controlled interrupted time series model
sort intervention time
*Respiratory tract infections
glm rti_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("RTI") replace
putexcel A1=matrix(r(table)), names
predict rti_yhat
gen rti_rate=rti_events/list_size
gen rti_pred_rate=rti_yhat/list_size
graph twoway (line rti_pred_rate time if intervention==1, lcolor(maroon)) (line rti_pred_rate time if intervention==0, lcolor(navy)) (scatter rti_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter rti_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/rti_plot.pdf", as(pdf)
*Acute respiratory tract infections
glm arti_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("aRTI") modify
putexcel A1=matrix(r(table)), names
predict arti_yhat
gen arti_rate=arti_events/list_size
gen arti_pred_rate=arti_yhat/list_size
graph twoway (line arti_pred_rate time if intervention==1, lcolor(maroon)) (line arti_pred_rate time if intervention==0, lcolor(navy)) (scatter arti_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter arti_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/arti_plot.pdf", as(pdf)
*Gastro intestinal infections
glm gastro_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("gastro") modify
putexcel A1=matrix(r(table)), names
predict gastro_yhat
gen gastro_rate=gastro_events/list_size
gen gastro_pred_rate=gastro_yhat/list_size
graph twoway (line gastro_pred_rate time if intervention==1, lcolor(maroon)) (line gastro_pred_rate time if intervention==0, lcolor(navy)) (scatter gastro_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter gastro_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/gastro_plot.pdf", as(pdf)
*Covid-19 diagnoses
glm coviddiag_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("CovidDiag") modify
putexcel A1=matrix(r(table)), names
predict coviddiag_yhat
gen coviddiag_rate=coviddiag_events/list_size
gen coviddiag_pred_rate=coviddiag_yhat/list_size
graph twoway (line coviddiag_pred_rate time if intervention==1, lcolor(maroon)) (line coviddiag_pred_rate time if intervention==0, lcolor(navy)) (scatter coviddiag_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter coviddiag_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/coviddiag_plot.pdf", as(pdf)
*Covid-19 symptoms - senstivie list
glm covidsympsens_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("CovidSympSens") modify
putexcel A1=matrix(r(table)), names
predict covidsympsens_yhat
gen covidsympsens_rate=covidsympsens_events/list_size
gen covidsympsens_pred_rate=covidsympsens_yhat/list_size
graph twoway (line covidsympsens_pred_rate time if intervention==1, lcolor(maroon)) (line covidsympsens_pred_rate time if intervention==0, lcolor(navy)) (scatter covidsympsens_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter covidsympsens_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/covidsympsens_plot.pdf", as(pdf)
*Covid-19 symptoms - specific list
glm covidsympspec_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("CovidSympSpec") modify
putexcel A1=matrix(r(table)), names
predict covidsympspec_yhat
gen covidsympspec_rate=covidsympspec_events/list_size
gen covidsympspec_pred_rate=covidsympspec_yhat/list_size
graph twoway (line covidsympspec_pred_rate time if intervention==1, lcolor(maroon)) (line covidsympspec_pred_rate time if intervention==0, lcolor(navy)) (scatter covidsympspec_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter covidsympspec_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/covidsympspec_plot.pdf", as(pdf)
*Antibiotics
glm antibio_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("antibio") modify
putexcel A1=matrix(r(table)), names
predict antibio_yhat
gen antibio_rate=antibio_events/list_size
gen antibio_pred_rate=antibio_yhat/list_size
graph twoway (line antibio_pred_rate time if intervention==1, lcolor(maroon)) (line antibio_pred_rate time if intervention==0, lcolor(navy)) (scatter antibio_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter antibio_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/antibio_plot.pdf", as(pdf)
*Hospital admissions
glm adm_events time intervention period time_intervention time_period intervention_period time_int_period, family(nb ml) link(log) exposure(population)
putexcel set "$gd/output/CITS_weekly.xlsx", sheet("adm") modify
putexcel A1=matrix(r(table)), names
predict adm_yhat
gen adm_rate=adm_events/list_size
gen adm_pred_rate=adm_yhat/list_size
graph twoway (line adm_pred_rate time if intervention==1, lcolor(maroon)) (line adm_pred_rate time if intervention==0, lcolor(navy)) (scatter adm_rate time if intervention==1, mcolor(maroon) msymbol(o)) (scatter adm_rate time if intervention==0, mcolor(navy) msymbol(o)), legend(order(1 "Intervention estimate" 2 "Control estimate" 3 "Intervention rates" 4 "Control rates")) xline(0, lcolor(black) lpattern(dash))
graph export "$gd/output/adm_plot.pdf", as(pdf)
** B. Practice level weekly time-series analysis
use "$gd/output/practice_weekly.dta", replace
*add practice covariates
merge m:1 practice_id using "$gd/output/practice_variables.dta"
drop _merge
gen time=week_date - 17.5
*Check smallest count
tabstat gastro_events list_size, stat(sum) by(intervention)
*create interaction variables
gen time_period = 0
replace time_period = time if period==1
gen time_intervention = 0
replace time_intervention = time if intervention==1
gen intervention_period = 0
replace intervention_period = 1 if intervention==1 & period==1
gen time_int_period = 0
replace time_int_period = time if intervention_period==1
*drop if practice_id==0
*export excel using "$gd/output/practice_weekly_glmm.xlsx", firstrow(variables) replace
*Random intercepts CITS
gen month=month(date)
menbreg rti_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("RTI") replace
putexcel A1=matrix(r(table)), names
menbreg arti_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("aRTI") modify
putexcel A1=matrix(r(table)), names
menbreg gastro_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("gastro") modify
putexcel A1=matrix(r(table)), names
menbreg coviddiag_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("CovidDiag") modify
putexcel A1=matrix(r(table)), names
menbreg covidsympsens_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("CovidSympSens") modify
putexcel A1=matrix(r(table)), names
menbreg covidsympspec_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("CovidSympSpec") modify
putexcel A1=matrix(r(table)), names
menbreg antibio_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("antibio") modify
putexcel A1=matrix(r(table)), names
menbreg adm_events time intervention period time_intervention time_period intervention_period time_int_period median_age deprivation_pctile female_pct i.month, exposure(population) || practice_id:
putexcel set "$gd/output/CITS_RandInt.xlsx", sheet("adm") modify
putexcel A1=matrix(r(table)), names