Am J Cardiovasc Drugs DOI 10.1007/s40256-014-0088-x
ORIGINAL RESEARCH ARTICLE
Patterns in the Use of Low-Dose Acetylsalicylic Acid and Other Therapies Following Upper Gastrointestinal Bleeding Marı´a E. Sa´ez • Antonio Gonza´lez-Pe´rez • Saga Johansson • Pe´ter Nagy • Luis A. Garcı´a Rodrı´guez
Ó Springer International Publishing Switzerland 2014
Abstract Background Anticoagulants and/or antiplatelet agents such as acetylsalicylic acid (ASA) are important in prevention of cardiovascular (CV) events, but may be associated with upper gastrointestinal bleeding (UGIB). However, discontinuing these agents may leave patients at risk of CV events. Objectives This study aimed to assess patterns of therapy after UGIB in routine clinical practice. Methods The Health Improvement Network UK primary care database was used to identify a cohort of patients aged 40–84 years with a UGIB event between 2000 and 2007 (n = 2,036). Patients were followed up for 1 year from the recorded UGIB. Re-prescription rates for antithrombotics and drugs that can modify the risk of UGIB were estimated at 30, 90, 180, and 365 days. Results At 365 days, the re-prescription rate was 43 % for ASA, 66 % for warfarin, 69 % for clopidogrel, and 49 % for dipyridamole. The re-prescription rate of gastroprotective agents at 365 days for current users of histamine H2-receptor antagonists was 36 % and that of proton pump inhibitors (PPIs) was 97 %. In patients who
Electronic supplementary material The online version of this article (doi:10.1007/s40256-014-0088-x) contains supplementary material, which is available to authorized users. M. E. Sa´ez A. Gonza´lez-Pe´rez L. A. Garcı´a Rodrı´guez (&) Spanish Centre for Pharmacoepidemiologic Research (CEIFE), c/Almirante 28, 2°, 28004 Madrid, Spain e-mail: [email protected]
were prescribed ASA before UGIB (n = 572), only 24 % were prescribed a PPI in the previous year. In patients who were prescribed ASA in the year after UGIB (n = 337), 92 % were prescribed a PPI. Conclusions Antiplatelet use fell after UGIB events. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics.
Key Points Use of antiplatelet agents fell after upper gastrointestinal bleeding (UGIB) events. In patients prescribed acetylsalicylic acid before a UGIB event, only 24 % had received a proton pump inhibitor (PPI) prescription in the previous year. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics. The concomitant use of gastroprotective agents to minimize gastrointestinal (GI) complications in patients at high GI risk and receiving antiplatelet agents could help patients remain on antiplatet treatment for continued prevention of cardiovascular events.
M. E. Sa´ez A. Gonza´lez-Pe´rez Andalusian Bioinformatics Research Center (CAEBi), Seville, Spain
1 Introduction
S. Johansson P. Nagy AstraZeneca R&D, Mo¨lndal, Sweden
The importance of antiplatelet agents, such as acetylsalicylic acid (ASA), in the prevention of cardiovascular (CV)
M. E. Sa´ez et al.
events is well established [1–4]; the use of low-dose ASA and clopidogrel, either as monotherapies or in combination, is currently a mainstay of therapy in the secondary prevention of CV events [5–8]. However, gastrointestinal (GI) adverse effects such as ulceration and upper GI bleeding (UGIB) have been associated with the use of antiplatelet agents, including low-dose ASA [9, 10]. Factors associated with an increased risk of ASA-related upper GI complications include advanced age, a previous history of ulcer or GI bleeding, ASA dose, and concomitant therapy with a nonsteroidal anti-inflammatory drug (NSAID), corticosteroid, anticoagulant, or additional antiplatelet drug [10–15]. NSAIDs are widely used medicines for the treatment of arthritis and many other inflammatory conditions. With high doses and prolonged use, it is likely that more serious adverse effects such as GI ulceration or bleeding may occur [16, 17]. The aim of this study was to assess the changes in use of antithrombotics (low-dose ASA and other therapies, including clopidogrel, dipyridamole, and warfarin) and drugs that can modify the risk of UGIB (such as NSAIDs [including traditional NSAIDs (tNSAIDs)] and selective inhibitors of cyclooxygenase-2 [coxibs], proton pump inhibitors [PPIs], histamine H2-receptor antagonists [H2RAs], steroids, and selective serotonin reuptake inhibitors [SSRIs]) before, immediately after, and 1 year after an episode of UGIB in routine clinical practice.
2 Methods A retrospective cohort study was conducted using individuals identified from The Health Improvement Network (THIN) primary care database in the UK. THIN contains anonymized data on almost 4 million patients, systematically recorded by primary care physicians (PCPs) as part of routine patient care and sent to THIN for use in research projects. The validity of THIN in pharmacoepidemiological research has been supported by evidence from previous studies, particularly when UGIB is the outcome of interest [18, 19]. In this analysis, the study population included patients with UGIB identified from a previous study [10] who had available follow-up data. Briefly, THIN was used to identify patients aged 40–84 years with a diagnosis of nonvariceal UGIB, from 1 January 2000 to 31 December 2007 (n = 2,036). The study was approved by a multicenter research ethics committee (National Health Service Research Ethics Committee reference number: 05/MRE02/70). Individuals were included in the source population only if they had been enrolled with their PCP for at least 2 years and had a computerized prescription history of at least 1 year. A number of exclusion criteria were used in order
to improve inference about causality of any association found in this study [10]. Patients in the study cohort were followed up from their start date (initial date of recorded UGIB event) for 1 year. Current use of each study drug (low-dose ASA, clopidogrel, dipyridamole, warfarin, NSAIDs [tNSAIDs and coxibs], PPIs, H2RAs, steroids, and SSRIs) was defined as having a recorded prescription that lasted until, or ended in the 7 days before, the start date. Chronic use was defined as current use with at least 1 year of continuous prescription history before the start date. 2.1 Statistical Analysis Re-prescription rates (and 95 % confidence intervals [CIs]) for each drug were estimated at 30, 90, 180, and 365 days after the start date using the Kaplan–Meier method, and reported as the proportion of current users of each drug who received a new prescription during the designated time interval. Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole, and antithrombotics comprised antiplatelet agents and warfarin. Re-prescription rates were also stratified by age, sex, and calendar year. Antiplatelet users were classified according to the indication at first use. Cox proportional hazards models, adjusted for age, sex, and calendar year, were used to analyze the effects of antiplatelet indication and use of other medications on represcription rates for each drug, including before and after receiving a prescription for a PPI, using models with timedependent variables. An additional analysis identified ASA initiators among non-ASA users in the year prior to the UGIB. In order to compare this rate of ASA initiation with that of an otherwise similar group of individuals who had not had a UGIB, a comparison group was identified from the study cohort that gave rise to the UGIB cases. Thus, for each one of the 1,306 patients who had not used ASA in the year before the UGIB, we selected a control from the comparison group (date-, age- and sex-matched, and who had not used ASA in the previous year). A Cox proportional hazards model was used to obtain estimates of the likelihood of initiating ASA in both groups, adjusted for prior comorbidity (ischemic stroke, transient ischemic attack, ischemic heart disease, diabetes mellitus, and hypertension).
3 Results Overall, for current users, the cumulative re-prescription rate of every study drug increased across each time point of follow-up (30, 90, 180, and 365 days), although the increase was greater for some drugs than for others.
Patterns in the Use of Low-Dose Acetylsalicylic Acid
3.1 Antithrombotics Re-prescription rates in current users were lower at each time point for ASA than for other antithrombotics; at 365 days, the re-prescription rate for ASA was 43 %, compared with 66 % for warfarin, 69 % for clopidogrel, and 49 % for dipyridamole (Table 1). Re-prescription rates for current users were lower at each time point for ASA than for clopidogrel or warfarin (Table 1). When analyzed as a group, the re-prescription rate of antithrombotics in current users at 365 days was 65 % and the re-prescription rate of antiplatelet agents was 61 % (Table 1). Although only 43 % of current ASA users were represcribed ASA during follow-up, 27 % (90/336) of those who were not re-prescribed ASA were prescribed a different antiplatelet agent. Most of these patients (79/90; 88 %) switched from ASA to clopidogrel. Similarly, only 49 % of dipyridamole users were re-prescribed this drug after UGIB, but 54 % (7/13) of those who were not represcribed dipyridamole were prescribed a different antiplatelet agent, with most (6/7; 86 %) switching to clopidogrel. Switching was less frequent among clopidogrel users; only two patients switched to ASA after UGIB. Some warfarin users also switched to antiplatelet agents. Among the 43 warfarin users who were not re-prescribed the drug, 37 % (n = 16) received a prescription for an antiplatelet agent, mainly clopidogrel (n = 8) or ASA (n = 7). Re-prescription rates among a subset of chronic users of antithrombotics ([1 year of continuous prescription history
before start date) were similar to those among all current users. Cumulative re-prescription rates at 365 days were as follows: ASA 46 %; warfarin 66 %; clopidogrel 71 %; and dipyridamole 56 % (see the Electronic Supplementary Material [ESM] online resource 1). Similar re-prescription rates for antithrombotic agents were observed in men and women (Fig. 1a). When analyzed by age, there was a trend for lower ASA and overall antithrombotic re-prescription rates in older patients than in younger patients 365 days after UGIB. Re-prescription rates for ASA were 58 %, 44 %, and 42 % in patients aged 40–54, 55–69, and 70–85 years, respectively (Fig. 1b). A similar trend was observed when analyzing antiplatelet agents as a group; this was statistically significant (ptrend = 0.01). When analyzed by calendar year, ASA re-prescription rates after UGIB events increased during the study period from 37 % in 2000 to 48 % in 2007 (ESM 2), but this trend was not statistically significant (ptrend = 0.21). Compared with patients initiating antiplatelet therapy for primary prevention of CV events, those who initiated antiplatelet therapy after a myocardial infarction or an episode of unstable angina had a higher probability of continuing antiplatelet therapy after UGIB (hazard ratio [HR] for continuing antiplatelet therapy after myocardial infarction 1.81 [95 % CI 1.31–2.52]; unstable angina 2.63 [95 % CI 1.32–5.24]) (Table 2). Compared with patients initiating ASA therapy for primary prevention of CV events, those who initiated ASA after a myocardial infarction also had a higher probability of continuing to receive ASA after UGIB (HR 1.70 [95 % CI 1.13–2.55]),
Table 1 Re-prescription rates in current users after upper gastrointestinal bleeding, stratified by time of follow-up Drug
Patients (n)
Re-prescription rate (95 % CI) 30 days
90 days
180 days
365 days
ASA
572
0.22 (0.19–0.26)
0.35 (0.31–0.39)
0.41 (0.37–0.45)
0.43 (0.39–0.47)
Warfarin
113
0.27 (0.19–0.36)
0.49 (0.40–0.59)
0.62 (0.53–0.71)
0.66 (0.57–0.75)
61 24
0.29 (0.19–0.42) 0.35 (0.19–0.58)
0.53 (0.41–0.66) 0.40 (0.23–0.63)
0.66 (0.54–0.78) 0.49 (0.31–0.71)
0.69 (0.57–0.81) 0.49 (0.31–0.71)
Antiplatelet agentsa
614
0.31 (0.28–0.35)
0.49 (0.45–0.53)
0.57 (0.53–0.61)
0.61 (0.57–0.64)
Antithromboticsb
706
0.32 (0.28–0.35)
0.51 (0.47–0.55)
0.60 (0.57–0.64)
0.65 (0.61–0.68)
Clopidogrel Dipyridamole
PPIs
276
0.67 (0.61–0.72)
0.92 (0.89–0.95)
0.95 (0.92–0.97)
0.97 (0.95–0.99)
H2RAs
103
0.17 (0.11–0.26)
0.25 (0.18–0.35)
0.34 (0.26–0.44)
0.36 (0.28–0.46)
SSRIs
97
0.51 (0.42–0.62)
0.86 (0.78–0.92)
0.91 (0.85–0.96)
0.91 (0.85–0.96)
Steroids
70
0.31 (0.21–0.44)
0.58 (0.46–0.70)
0.68 (0.56–0.79)
0.73 (0.62–0.84)
NSAIDs
347
0.09 (0.06–0.13)
0.15 (0.12–0.20)
0.20 (0.16–0.24)
0.24 (0.20–0.29)
tNSAIDs
295
0.08 (0.06–0.12)
0.12 (0.09–0.17)
0.15 (0.11–0.20)
0.18 (0.14–0.24)
56
0.02 (0.00–0.13)
0.06 (0.02–0.17)
0.11 (0.05–0.23)
0.13 (0.07–0.26)
Coxibs
ASA acetylsalicylic acid, CI confidence interval, coxib cyclooxygenase-2 selective inhibitor, H2RA histamine H2-receptor antagonist, NSAID nonsteroidal anti-inflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor, tNSAID traditional NSAID a
Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole
b
Antithrombotics comprised antiplatelet agents and warfarin
M. E. Sa´ez et al.
a
100
98 97
97
Male Female
89
Re-prescription rate (%)
Fig. 1 Re-prescription rates of study drugs 365 days after upper gastrointestinal bleeding, stratified by (a) sex and (b) age. *Antiplatelet agents comprised acetylsalicylic acid, clopidogrel, and dipyridamole. Antithrombotics comprised antiplatelet agents and warfarin. ASA acetylsalicylic acid, coxib cyclooxygenase-2 selective inhibitor, H2RA histamine H2-receptor antagonist, N/A data not available, NSAID nonsteroidal anti-inflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor, tNSAID traditional NSAID
80
80 73
72 67
65
62
64
60 61
60
66
64
45
42
40
40
32 26
24 24 20
20
19
17 7
0 A
AS
b
n
ari
rf Wa
rel
ole
og
pid
Clo
am
rid
y Dip
tics
ets
tel
pla
ti An
*
†
Is
PP
bo
om
r tith
s
RA H2
RIs
SS
s
ds
AID
roi
Ste
NS
s
98 98
97
94
40–54 years 55–69 years 70–85 years
90 86
Re-prescription rate (%)
85
76
74
72 66
65
60
71
65
68
64
62
Co
An
100
80
s
xib
AID
tNS
63 58
58 50 44
50
47
42
39 39
40
28
25
26
21
20
16
N/A
0 A AS
Wa
N/A
l
n
ri rfa
re og
id
p Clo
da
ri ipy
D
N/A
cs oti
s
le mo
la
p nti
*A
et tel †
b
om thr
Is PP
RA 2
S
ti
An
14
N/A
Is SR
s
H
19
18
s AID
s
St
id ero
NS
s AID
tNS
s
xib
Co
Table 2 Antiplatelet indication and re-prescription rates 365 days after upper gastrointestinal bleeding Indication
Patients (n)
Re-prescription rate (95 % CI)
HR (95 % CI)a
Primary prevention
169 (27.5 %)
0.53 (0.46–0.61)
1.00
Peripheral vascular disease
93 (15.2 %)
0.49 (0.39–0.60)
0.92 (0.64–1.32)
Ischemic heart disease
145 (23.6 %)
0.61 (0.53–0.69)
1.25 (0.93–1.69)
Cerebrovascular disease
106 (17.3 %)
0.67 (0.58–0.76)
1.49 (1.08–2.05)
Unstable angina
12 (1.9 %)
0.79 (0.52–0.96)
2.63 (1.32–5.24)
Myocardial infarction
89 (14.5 %)
0.74 (0.65–0.83)
1.81 (1.31–2.52)
CI confidence interval, HR hazard ratio a Estimates obtained using a Cox proportional hazards model that included age, sex, calendar year, and antiplatelet indication
although the absolute number of ASA re-prescribing was noticeably lower than for all antiplatelets (ESM 3). 3.2 Initiation of Low-dose Acetylsalicylic Acid in the First Year after Upper Gastrointestinal Bleeding We found that 50 of the 1,306 patients with UGIB who had not used ASA in the previous year initiated ASA treatment
during the first year after the UGIB, corresponding to a prescribing rate of 4.0 % (95 % CI 3.1–5.3). During this period after the UGIB, one or more PPI prescriptions were issued in 92 % of these ASA initiators. We found that 99 of the comparison cohort (1,306 date-, age-, and sex-matched controls, who had not used ASA in the previous year) initiated ASA after 1 year of follow-up, corresponding to a prescribing rate of 7.5 % (95 % CI 6.2–9.1). During this period, one or more PPI prescriptions
Patterns in the Use of Low-Dose Acetylsalicylic Acid
were issued in 20 % of these ASA initiators. Overall, initiating ASA treatment was less frequent during the first year after a UGIB (HR 0.53 [95 % CI 0.38–0.75]) than in the comparison cohort without UGIB. The estimate was similar after adjustment for prior comorbidity (HR 0.46 [95 % CI 0.33–0.65]). 3.3 Gastroprotective Agents The cumulative re-prescription rates of gastroprotective agents 1 year after UGIB for current users of H2RAs was 36 %, and for PPIs it was 97 % (Table 1). The lower rate of H2RA re-prescription may be explained by switching to PPIs: 94 % (63/67) of current users of H2RAs who were not re-prescribed H2RAs received a PPI. Among chronic users of gastroprotective agents, re-prescription rates were similar to those of current users overall (ESM 1). Cumulative re-prescription rates at 365 days were 43 % and 99 % for H2RAs and PPIs, respectively. Similar re-prescription rates for H2RAs and PPIs were observed in men and women (Fig. 1a). When analyzed by age, PPI re-prescription rates 1 year after UGIB were similar in all age groups (94 %, 98 %, and 98 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). When analyzed by calendar year, the PPI represcription rate after UGIB was 97 % in 2002 and remained at 97 % in 2007 (ESM 2). Re-prescription of H2RAs increased from 43 % in 2000 to 71 % in 2001, but decreased in subsequent years to 20 % in 2007; this decrease was statistically significant (ptrend = 0.011). 3.4 Anti-Inflammatory Drugs Among the anti-inflammatory drugs studied, the highest represcription rate was for steroids, with 73 % of patients being re-prescribed treatment in the 1-year period after UGIB. In contrast, at 1 year after UGIB, the re-prescription rates for NSAIDs were the lowest among all the drugs analyzed, at 24 %, and were lower for coxibs than for tNSAIDs (Table 1). Among current users of tNSAIDs who did not receive a new prescription, 9 % (20/224) switched to coxibs after UGIB. Among chronic users ([1 year of continuous prescription history before the start date), re-prescription rates of coxibs and tNSAIDs were similar, but re-prescription rates for both were greater than for current users. Cumulative re-prescription rates at 365 days for chronic users were as follows: NSAIDs 38 %; tNSAIDs 29 %; and coxibs 31 % (ESM 1). Similar re-prescription rates for NSAIDS overall and tNSAIDs were observed in men and women (Fig. 1a). When analyzed by age, NSAID re-prescription rates 1 year after UGIB were similar in all age groups (re-prescription
rates were 28 %, 21 %, and 25 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). Steroids were more often re-prescribed to older patients; cumulative re-prescription rates were 76 % for patients aged 70–85 years compared with 50 % for patients aged 40–54 years (no statistically significant difference). When analyzed by calendar year, NSAID re-prescription rates after UGIB were low at 19 % in 2000 and remained low at 20 % in 2007 (ESM 2). Re-prescription of steroids fluctuated, peaking at 91 % in 2002, but falling to 50 % in 2007. 3.5 Selective Serotonin Reuptake Inhibitors For current SSRI users, the re-prescription rate 1 year after UGIB was 91 % (Table 1). Among a subset of chronic SSRI users, re-prescription rates were similar to overall current users at 95 % (ESM 1). SSRIs were re-prescribed in 97 % of men and 89 % of women (Fig. 1a). When analyzed by age, SSRI re-prescription rates 1 year after UGIB were similar in all age groups (86 %, 97 %, and 90 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). When analyzed by calendar year, SSRI re-prescription rates after UGIB were generally high, at 71 % in 2000 and 80 % in 2007 (ESM 2). 3.6 Concomitant Use of Proton Pump Inhibitors and Antithrombotics In patients who were prescribed ASA before UGIB (n = 572), only 24 % had received a PPI prescription in the previous year. In patients who were prescribed ASA in the year after UGIB (n = 337), 92 % were prescribed a PPI. Patients who were taking both PPIs and clopidogrel at the time of UGIB had a reduced probability of being represcribed clopidogrel after UGIB compared with patients who used only clopidogrel at the time of UGIB (HR 0.20 [95 % CI 0.06–0.70]) (Table 3). In patients who were taking concomitant PPIs with ASA, warfarin, steroids, or NSAIDs at the time of UGIB, there was a trend towards increased re-prescription of the drug after UGIB compared with patients who used only ASA, warfarin, steroids, or NSAIDs, although this was not statistically significant (HR for ASA 1.26 [95 % CI 0.88–1.81]; warfarin 1.16 [95 % CI 0.55–2.43]; steroids 1.39 [95 % CI 0.68–2.81]; NSAIDs 1.24 [95 % CI 0.66–2.34]) (Table 3). We found that, once patients were prescribed a PPI after UGIB, the overall likelihood of being re-prescribed antithrombotics was increased (HR 2.49 [95 % CI 1.92–3.24]), antiplatelet agents overall (HR 2.38 [95 % CI 1.79–3.18]), ASA (HR 1.68 [95 % CI 1.19–2.37]), and SSRIs (HR 2.33 [95 % CI 1.40–3.89]). For NSAIDs (HR 1.74 [95 % CI
M. E. Sa´ez et al. Table 3 Re-prescription rates and hazard ratios, stratified by concomitant use of proton pump inhibitors at the time of upper gastrointestinal bleeding
Drug
Patients (n)
Patients with re-prescription at 365 days (n)
Re-prescription rate (95 % CI)
HR (95 % CI), adjusted for age, sex, and calendar year
437
171
0.41 (0.37–0.46)
1.00
78
37
0.49 (0.39–0.61)
1.26 (0.88–1.81)
Non-PPI users at start date
41
29
0.74 (0.60–0.86)
1.00
PPI users at start date
11
3
0.30 (0.11–0.67)
0.20 (0.06–0.70)
93
57
0.66 (0.55–0.75)
1.00
12
8
0.71 (0.44–0.93)
1.16 (0.55–2.43)
465
264
0.60 (0.55–0.64)
1.00
85
47
0.58 (0.47–0.69)
0.99 (0.72–1.35)
541
333
0.65 (0.61–0.69)
1.00
96
56
0.61 (0.51–0.71)
0.95 (0.71–1.26)
Non-PPI users at start date
63
55
0.92 (0.83–0.97)
1.00
PPI users at start date
22
20
0.91 (0.75–0.98)
1.00 (0.60–1.68)
Non-PPI users at start date
48
29
0.69 (0.55–0.83)
1.00
PPI users at start date
15
11
0.81 (0.56–0.96)
1.39 (0.68–2.81)
ASA Non-PPI users at start date PPI users at start date Clopidogrela
Warfarin Non-PPI users at start date PPI users at start date Antiplatelet agentsb Non-PPI users at start date PPI users at start date ASA acetylsalicylic acid, CI confidence interval, HR hazard ratio, NSAID nonsteroidal antiinflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor a
At the start date, approximately one-third of clopidogrel users were concomitant users of ASA
b
Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole
c
Antithrombotics comprised antiplatelet agents and warfarin
Antithromboticsc Non-PPI users at start date PPI users at start date SSRIs
Steroids
NSAIDs Non-PPI users at start date PPI users at start date
0.96–3.17]), steroids (HR 1.96 [95 % CI 0.89–4.34]), and warfarin (HR 1.79 [95 % CI 0.93–3.43]), the increased represcription did not reach statistical significance.
4 Discussion This study took place between 2000 and 2007 and reports prescribing practices for low-dose ASA and other therapies after UGIB. As expected, the results showed that antithrombotic use fell after UGIB. ASA re-prescription after UGIB increased slightly over the study period, but fewer than half (43 %) of patients receiving ASA at the time of UGIB received a new prescription in the year following UGIB. Likewise, initiating ASA therapy was half as frequent among individuals with a recent episode of UGIB as among individuals free of UGIB. Re-prescription rates were higher among users of other antithrombotics, including clopidogrel and warfarin, than among users of ASA (although based on small numbers of patients). These re-prescription rates are in line with those reported in a recent study in Denmark, which also found a reduction in re-prescription rates of antithrombotics after
272
57
0.22 (0.18–0.28)
1.00
47
12
0.28 (0.17–0.44)
1.24 (0.66–2.34)
UGIB [20]. This has important clinical implications, since a reduction in antithrombotic use after UGIB would result in an increase in CV risk in these patients [21]. Antithrombotic therapy re-prescription rates decreased with increasing age, indicating caution in re-prescribing such agents in older patients who have had UGIB, despite an increased CV risk in these individuals [22]. Among antiinflammatory drugs, the highest re-prescription rate was observed for steroids. NSAIDs had the lowest represcription rate of the drugs studied, which may be a reflection of the perceived low benefit–risk ratio for these drugs. Although re-prescription of ASA was only 43 % at 1 year after UGIB, the re-prescription of antiplatelet agents overall was 61 %, indicating that many individuals continue or switch antiplatelet agents after UGIB. Among the gastroprotective drugs, the re-prescription rate for PPI users was 97 %, whereas it was only 36 % for H2RA users. In patients who were prescribed ASA before UGIB, only one-quarter had received a PPI prescription in the previous year, but in patients re-prescribed ASA in the year after UGIB, nearly all (92 %) were prescribed a PPI. Incidentally, concomitant use of PPI was identical among
Patterns in the Use of Low-Dose Acetylsalicylic Acid
ASA initiators after UGIB. This reflects guidelines for patients receiving antiplatelet agents (such as ASA), which recommend the use of PPIs for patients who have a history of UGIB, or who have multiple risk factors for bleeding, including combined use of antiplatelet agents and NSAIDs [14, 15]. For patients receiving dual antiplatelet therapy, recent guidelines do not prohibit the use of PPIs in appropriate clinical settings, but highlight the potential risks and benefits from the use of PPIs in combination with clopidogrel [8]. Analysis of the re-prescription rates of the drugs studied in patients who used concomitant PPIs at the time of UGIB indicated that, for concomitant use of PPIs with ASA, warfarin, steroids, or NSAIDs, there was an increased likelihood of re-prescription after UGIB compared with patients who used only ASA, warfarin, steroids, or NSAIDs. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics overall, and for the individual study drugs ASA, warfarin, and SSRIs. This is in line with results from other studies, which confirm the effectiveness of PPIs in reducing re-bleeding [23–26]. The current study was undertaken using THIN, a database encompassing a large, representative sample of the UK population, which has been validated for use in epidemiological studies [19], therefore giving our results good external validity. Although over-the-counter low-dose ASA is not recorded in this population, cross-sectional studies have shown that use of low-dose ASA for secondary prevention is commonly prescription based in the UK [27, 28], with patients aged over 60 years receiving free prescriptions, reducing the likelihood of over-the-counter medication use. Some studies have reported rates of overthe-counter low-dose ASA use in secondary prevention populations of between 20 and 40 % [27, 28]; however, more recent data from our own manual review of free-text comments in a random sample of 500 individuals in THIN (aged [40 years, attending primary care services, and without any low-dose ASA prescription) indicated that as few as 5 % of individuals had a mention of overthe-counter low-dose ASA use in their clinical records. As with ASA therapy, some NSAIDs and PPIs are available over the counter and, because such use is not recorded by PCPs, this is a potential limitation of the study, although such practices are likely to be infrequent immediately after an episode of UGIB. In addition, patients leaving the hospital are likely to be provided with medication by the hospital for at least a month. Thus re-prescription rates within 30 days may be missing these hospital prescriptions, although, partly for this reason, the 30-day estimate was not the primary endpoint in the current study. ASA is recommended for long-term maintenance treatment in patients who have a history of CV events [5–8]. It
is reassuring that, in this study, patients who initiated antiplatelet therapy following a myocardial infarction or an episode of unstable angina had a greater probability of continuing antiplatelet therapy after UGIB than those initiating antiplatelet therapy for primary prevention. However, it is clear that not all patients receiving ASA for secondary prevention (who should be receiving continuous long-term ASA) are being re-prescribed antiplatelet agents after UGIB. Discontinuation of ASA has been shown to increase the probability of having a CV event, particularly in the first few days and weeks after stopping the medication [29–31]. We have previously shown a 43 % increase in the risk of CV death or nonfatal myocardial infarction and a 40 % increase in the risk of ischemic stroke or transient ischemic attack among recent low-dose ASA discontinuers [29, 32]. Patients who are not re-prescribed therapy may be left at high risk of secondary CV events if they are not treated with an alternative antithrombotic medication. Several recent studies have indicated that some patients may discontinue their ASA owing to GI adverse effects [33–35]. Assessment of both GI and CV risks versus the benefits of low-dose ASA for individual patients can sometimes be difficult in clinical practice. A recent paper by Lanas et al. [36] reports the development of a risk calculator based on the incidence of CV events and upper GI complications according to the presence of different risk factors. This may prove to be a useful tool to help physicians estimate CV and GI risks and to facilitate the clinical decisionmaking process. Acknowledgments The study was funded in part by AstraZeneca R&D, Mo¨lndal, Sweden. Writing support was provided by Dr Rowan Pearce of Oxford PharmaGenesisTM Ltd and funded by AstraZeneca R&D, Mo¨lndal, Sweden. Conflict of interest statement Marı´a E. Sa´ez, Antonio Gonza´lezPe´rez, and Luis A. Garcı´a Rodrı´guez are employees of the Spanish Centre for Pharmacoepidemiologic Research (CEIFE), which has received research funding from AstraZeneca R&D, Mo¨lndal, Sweden, and Bayer Pharma AG, Berlin, Germany. Luis A. Garcı´a Rodrı´guez has also served as a speaker and an advisory board member for AstraZeneca R&D, Mo¨lndal, Sweden, and Bayer Pharma AG, Berlin, Germany. Saga Johansson and Pe´ter Nagy are employees of AstraZeneca R&D, Mo¨lndal, Sweden.
References 1. Antiplatelet Trialists’ Collaboration. Collaborative overview of randomised trials of antiplatelet therapy—I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ. 1994;308:81–106. 2. Antithrombotic Trialists’ Collaboration. Collaborative metaanalysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ. 2002;324:71–86.
M. E. Sa´ez et al. 3. Patrono C, Coller B, Dalen JE, et al. Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest. 2001;119:39S–63S. 4. Patrono C, Coller B, Dalen JE, et al. Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest. 1998;114:470S–88S. 5. Perk J, De Backer G, Gohlke H, et al. European guidelines on cardiovascular disease prevention in clinical practice (version 2012): The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Eur Heart J. 2012;33:1635–701. 6. Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC). Eur Heart J. 2012;33: 2569–619. 7. O’Gara PT, Kushner FG, Ascheim DD, et al. ACCF/AHA Guideline for the Management of ST-Elevation Myocardial Infarction: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2013;127:e362–425. 8. Anderson JL, Adams CD, Antman EM, et al. 2012 ACCF/AHA focused update incorporated into the ACCF/AHA 2007 guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2013;127:e663–828. 9. Hernandez-Diaz S, Garcia Rodriguez LA. Cardioprotective aspirin users and their excess risk of upper gastrointestinal complications. BMC Med. 2006;4:22. 10. Garcı´a Rodrı´guez LA, Lin KJ, Herna´ndez-Dı´az S, et al. Risk of upper gastrointestinal bleeding with low-dose acetylsalicylic acid alone and in combination with clopidogrel and other medications. Circulation. 2011;123:1108–15. 11. Lanas A, Bajador E, Serrano P, et al. Nitrovasodilators, low-dose aspirin, other nonsteroidal antiinflammatory drugs, and the risk of upper gastrointestinal bleeding. N Engl J Med. 2000;343:834–9. 12. Lanas A, Garcı´a Rodrı´guez LA, Arroyo MT, et al. Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal antiinflammatory drugs, aspirin and combinations. Gut. 2006;55:1731–8. 13. Lanas A. Gastrointestinal bleeding associated with low-dose aspirin use: relevance and management in clinical practice. Expert Opin Drug Saf. 2011;10:45–54. 14. Bhatt DL, Scheiman J, Abraham NS, et al. ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. Circulation. 2008;118: 1894–909. 15. Abraham NS, Hlatky MA, Antman EM, et al. ACCF/ACG/AHA 2010 expert consensus document on the concomitant use of proton pump inhibitors and thienopyridines: a focused update of the ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use. Circulation. 2010;122:2619–33. 16. Garcı´a Rodrı´guez LA, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population. Gastroenterology. 2007; 132:498–506. 17. Herna´ndez-Dı´az S, Garcı´a Rodrı´guez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
tract bleeding/perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med. 2000;160:2093–9. Margulis AV, Garcı´a Rodrı´guez LA, Herna´ndez-Dı´az S. Positive predictive value of computerized medical records for uncomplicated and complicated upper gastrointestinal ulcer. Pharmacoepidemiol Drug Saf. 2009;18:900–9. Lewis JD, Schinnar R, Bilker WB, et al. Validation studies of the health improvement network (THIN) database for pharmacoepidemiology research. Pharmacoepidemiol Drug Saf. 2007;16:393–401. Dall M, Depont Christensen R, de Schaffalitzky Muckadell OB, et al. Re-prescribing of causative drugs in persons discharged after serious drug-induced upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2012;35:948–54. Sung JJ, Lau JY, Ching JY, et al. Continuation of low-dose aspirin therapy in peptic ulcer bleeding: a randomized trial. Ann Intern Med. 2010;152:1–9. Segal WN, Cello JP. Hemorrhage in the upper gastrointestinal tract in the older patient. Am J Gastroenterol. 1997;92:42–6. Lau JY, Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med. 2000;343:310–6. Sung JJ, Barkun A, Kuipers EJ, et al. Intravenous esomeprazole for prevention of recurrent peptic ulcer bleeding: a randomized trial. Ann Intern Med. 2009;150:455–64. Chan FK, Ching JY, Hung LC, et al. Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med. 2005;352:238–44. Masso´ Gonza´lez EL, Garcı´a Rodrı´guez LA. Proton pump inhibitors reduce the long-term risk of recurrent upper gastrointestinal bleeding: an observational study. Aliment Pharmacol Ther. 2008;28:629–37. Bedson J, Whitehurst T, Lewis M, et al. Factors affecting overthe-counter use of aspirin in the secondary prophylaxis of cardiovascular disease. Br J Gen Pract. 2001;51:1001–3. Hopper S, Pierce M. Aspirin after myocardial infarction: the importance of over-the-counter use. Fam Pract. 1998;15(Suppl 1):S10–3. Garcı´a Rodrı´guez LA, Cea-Soriano L, Martı´n-Merino E, et al. Discontinuation of low dose aspirin and risk of myocardial infarction: case–control study in UK primary care. BMJ. 2011;343:d4094. Ferrari E, Benhamou M, Cerboni P, et al. Coronary syndromes following aspirin withdrawal: a special risk for late stent thrombosis. J Am Coll Cardiol. 2005;45:456–9. Biondi-Zoccai GG, Lotrionte M, Agostoni P, et al. A systematic review and meta-analysis on the hazards of discontinuing or not adhering to aspirin among 50,279 patients at risk for coronary artery disease. Eur Heart J. 2006;27:2667–74. Garcı´a Rodrı´guez LA, Cea Soriano L, Hill C, et al. Increased risk of stroke after discontinuation of acetylsalicylic acid: a UK primary care study. Neurology. 2011;76:740–6. Martı´n-Merino E, Johansson S, Bueno H, et al. Discontinuation of low-dose acetylsalicylic acid therapy in UK primary care: incidence and predictors in patients with cardiovascular disease. Prag Obs Res. 2012;3:1–9. Pratt S, Thompson VJ, Elkin EP, et al. The impact of upper gastrointestinal symptoms on nonadherence to, and discontinuation of, low-dose acetylsalicylic acid in patients with cardiovascular risk. Am J Cardiovasc Drugs. 2010;10:281–8. Moberg C, Naesdal J, Svedberg L-E, et al. Impact of gastrointestinal problems on adherence to low-dose acetylsalicylic acid: a quantitative study in patients with cardiovascular risk. Patient. 2011;4:103–13. Lanas A, Polo-Tomas M, Casado-Arroyo R. The aspirin cardiovascular/gastrointestinal risk calculator—a tool to aid clinicians in practice. Aliment Pharmacol Ther. 2013;37:738–48.
ORIGINAL RESEARCH ARTICLE
Patterns in the Use of Low-Dose Acetylsalicylic Acid and Other Therapies Following Upper Gastrointestinal Bleeding Marı´a E. Sa´ez • Antonio Gonza´lez-Pe´rez • Saga Johansson • Pe´ter Nagy • Luis A. Garcı´a Rodrı´guez
Ó Springer International Publishing Switzerland 2014
Abstract Background Anticoagulants and/or antiplatelet agents such as acetylsalicylic acid (ASA) are important in prevention of cardiovascular (CV) events, but may be associated with upper gastrointestinal bleeding (UGIB). However, discontinuing these agents may leave patients at risk of CV events. Objectives This study aimed to assess patterns of therapy after UGIB in routine clinical practice. Methods The Health Improvement Network UK primary care database was used to identify a cohort of patients aged 40–84 years with a UGIB event between 2000 and 2007 (n = 2,036). Patients were followed up for 1 year from the recorded UGIB. Re-prescription rates for antithrombotics and drugs that can modify the risk of UGIB were estimated at 30, 90, 180, and 365 days. Results At 365 days, the re-prescription rate was 43 % for ASA, 66 % for warfarin, 69 % for clopidogrel, and 49 % for dipyridamole. The re-prescription rate of gastroprotective agents at 365 days for current users of histamine H2-receptor antagonists was 36 % and that of proton pump inhibitors (PPIs) was 97 %. In patients who
Electronic supplementary material The online version of this article (doi:10.1007/s40256-014-0088-x) contains supplementary material, which is available to authorized users. M. E. Sa´ez A. Gonza´lez-Pe´rez L. A. Garcı´a Rodrı´guez (&) Spanish Centre for Pharmacoepidemiologic Research (CEIFE), c/Almirante 28, 2°, 28004 Madrid, Spain e-mail: [email protected]
were prescribed ASA before UGIB (n = 572), only 24 % were prescribed a PPI in the previous year. In patients who were prescribed ASA in the year after UGIB (n = 337), 92 % were prescribed a PPI. Conclusions Antiplatelet use fell after UGIB events. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics.
Key Points Use of antiplatelet agents fell after upper gastrointestinal bleeding (UGIB) events. In patients prescribed acetylsalicylic acid before a UGIB event, only 24 % had received a proton pump inhibitor (PPI) prescription in the previous year. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics. The concomitant use of gastroprotective agents to minimize gastrointestinal (GI) complications in patients at high GI risk and receiving antiplatelet agents could help patients remain on antiplatet treatment for continued prevention of cardiovascular events.
M. E. Sa´ez A. Gonza´lez-Pe´rez Andalusian Bioinformatics Research Center (CAEBi), Seville, Spain
1 Introduction
S. Johansson P. Nagy AstraZeneca R&D, Mo¨lndal, Sweden
The importance of antiplatelet agents, such as acetylsalicylic acid (ASA), in the prevention of cardiovascular (CV)
M. E. Sa´ez et al.
events is well established [1–4]; the use of low-dose ASA and clopidogrel, either as monotherapies or in combination, is currently a mainstay of therapy in the secondary prevention of CV events [5–8]. However, gastrointestinal (GI) adverse effects such as ulceration and upper GI bleeding (UGIB) have been associated with the use of antiplatelet agents, including low-dose ASA [9, 10]. Factors associated with an increased risk of ASA-related upper GI complications include advanced age, a previous history of ulcer or GI bleeding, ASA dose, and concomitant therapy with a nonsteroidal anti-inflammatory drug (NSAID), corticosteroid, anticoagulant, or additional antiplatelet drug [10–15]. NSAIDs are widely used medicines for the treatment of arthritis and many other inflammatory conditions. With high doses and prolonged use, it is likely that more serious adverse effects such as GI ulceration or bleeding may occur [16, 17]. The aim of this study was to assess the changes in use of antithrombotics (low-dose ASA and other therapies, including clopidogrel, dipyridamole, and warfarin) and drugs that can modify the risk of UGIB (such as NSAIDs [including traditional NSAIDs (tNSAIDs)] and selective inhibitors of cyclooxygenase-2 [coxibs], proton pump inhibitors [PPIs], histamine H2-receptor antagonists [H2RAs], steroids, and selective serotonin reuptake inhibitors [SSRIs]) before, immediately after, and 1 year after an episode of UGIB in routine clinical practice.
2 Methods A retrospective cohort study was conducted using individuals identified from The Health Improvement Network (THIN) primary care database in the UK. THIN contains anonymized data on almost 4 million patients, systematically recorded by primary care physicians (PCPs) as part of routine patient care and sent to THIN for use in research projects. The validity of THIN in pharmacoepidemiological research has been supported by evidence from previous studies, particularly when UGIB is the outcome of interest [18, 19]. In this analysis, the study population included patients with UGIB identified from a previous study [10] who had available follow-up data. Briefly, THIN was used to identify patients aged 40–84 years with a diagnosis of nonvariceal UGIB, from 1 January 2000 to 31 December 2007 (n = 2,036). The study was approved by a multicenter research ethics committee (National Health Service Research Ethics Committee reference number: 05/MRE02/70). Individuals were included in the source population only if they had been enrolled with their PCP for at least 2 years and had a computerized prescription history of at least 1 year. A number of exclusion criteria were used in order
to improve inference about causality of any association found in this study [10]. Patients in the study cohort were followed up from their start date (initial date of recorded UGIB event) for 1 year. Current use of each study drug (low-dose ASA, clopidogrel, dipyridamole, warfarin, NSAIDs [tNSAIDs and coxibs], PPIs, H2RAs, steroids, and SSRIs) was defined as having a recorded prescription that lasted until, or ended in the 7 days before, the start date. Chronic use was defined as current use with at least 1 year of continuous prescription history before the start date. 2.1 Statistical Analysis Re-prescription rates (and 95 % confidence intervals [CIs]) for each drug were estimated at 30, 90, 180, and 365 days after the start date using the Kaplan–Meier method, and reported as the proportion of current users of each drug who received a new prescription during the designated time interval. Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole, and antithrombotics comprised antiplatelet agents and warfarin. Re-prescription rates were also stratified by age, sex, and calendar year. Antiplatelet users were classified according to the indication at first use. Cox proportional hazards models, adjusted for age, sex, and calendar year, were used to analyze the effects of antiplatelet indication and use of other medications on represcription rates for each drug, including before and after receiving a prescription for a PPI, using models with timedependent variables. An additional analysis identified ASA initiators among non-ASA users in the year prior to the UGIB. In order to compare this rate of ASA initiation with that of an otherwise similar group of individuals who had not had a UGIB, a comparison group was identified from the study cohort that gave rise to the UGIB cases. Thus, for each one of the 1,306 patients who had not used ASA in the year before the UGIB, we selected a control from the comparison group (date-, age- and sex-matched, and who had not used ASA in the previous year). A Cox proportional hazards model was used to obtain estimates of the likelihood of initiating ASA in both groups, adjusted for prior comorbidity (ischemic stroke, transient ischemic attack, ischemic heart disease, diabetes mellitus, and hypertension).
3 Results Overall, for current users, the cumulative re-prescription rate of every study drug increased across each time point of follow-up (30, 90, 180, and 365 days), although the increase was greater for some drugs than for others.
Patterns in the Use of Low-Dose Acetylsalicylic Acid
3.1 Antithrombotics Re-prescription rates in current users were lower at each time point for ASA than for other antithrombotics; at 365 days, the re-prescription rate for ASA was 43 %, compared with 66 % for warfarin, 69 % for clopidogrel, and 49 % for dipyridamole (Table 1). Re-prescription rates for current users were lower at each time point for ASA than for clopidogrel or warfarin (Table 1). When analyzed as a group, the re-prescription rate of antithrombotics in current users at 365 days was 65 % and the re-prescription rate of antiplatelet agents was 61 % (Table 1). Although only 43 % of current ASA users were represcribed ASA during follow-up, 27 % (90/336) of those who were not re-prescribed ASA were prescribed a different antiplatelet agent. Most of these patients (79/90; 88 %) switched from ASA to clopidogrel. Similarly, only 49 % of dipyridamole users were re-prescribed this drug after UGIB, but 54 % (7/13) of those who were not represcribed dipyridamole were prescribed a different antiplatelet agent, with most (6/7; 86 %) switching to clopidogrel. Switching was less frequent among clopidogrel users; only two patients switched to ASA after UGIB. Some warfarin users also switched to antiplatelet agents. Among the 43 warfarin users who were not re-prescribed the drug, 37 % (n = 16) received a prescription for an antiplatelet agent, mainly clopidogrel (n = 8) or ASA (n = 7). Re-prescription rates among a subset of chronic users of antithrombotics ([1 year of continuous prescription history
before start date) were similar to those among all current users. Cumulative re-prescription rates at 365 days were as follows: ASA 46 %; warfarin 66 %; clopidogrel 71 %; and dipyridamole 56 % (see the Electronic Supplementary Material [ESM] online resource 1). Similar re-prescription rates for antithrombotic agents were observed in men and women (Fig. 1a). When analyzed by age, there was a trend for lower ASA and overall antithrombotic re-prescription rates in older patients than in younger patients 365 days after UGIB. Re-prescription rates for ASA were 58 %, 44 %, and 42 % in patients aged 40–54, 55–69, and 70–85 years, respectively (Fig. 1b). A similar trend was observed when analyzing antiplatelet agents as a group; this was statistically significant (ptrend = 0.01). When analyzed by calendar year, ASA re-prescription rates after UGIB events increased during the study period from 37 % in 2000 to 48 % in 2007 (ESM 2), but this trend was not statistically significant (ptrend = 0.21). Compared with patients initiating antiplatelet therapy for primary prevention of CV events, those who initiated antiplatelet therapy after a myocardial infarction or an episode of unstable angina had a higher probability of continuing antiplatelet therapy after UGIB (hazard ratio [HR] for continuing antiplatelet therapy after myocardial infarction 1.81 [95 % CI 1.31–2.52]; unstable angina 2.63 [95 % CI 1.32–5.24]) (Table 2). Compared with patients initiating ASA therapy for primary prevention of CV events, those who initiated ASA after a myocardial infarction also had a higher probability of continuing to receive ASA after UGIB (HR 1.70 [95 % CI 1.13–2.55]),
Table 1 Re-prescription rates in current users after upper gastrointestinal bleeding, stratified by time of follow-up Drug
Patients (n)
Re-prescription rate (95 % CI) 30 days
90 days
180 days
365 days
ASA
572
0.22 (0.19–0.26)
0.35 (0.31–0.39)
0.41 (0.37–0.45)
0.43 (0.39–0.47)
Warfarin
113
0.27 (0.19–0.36)
0.49 (0.40–0.59)
0.62 (0.53–0.71)
0.66 (0.57–0.75)
61 24
0.29 (0.19–0.42) 0.35 (0.19–0.58)
0.53 (0.41–0.66) 0.40 (0.23–0.63)
0.66 (0.54–0.78) 0.49 (0.31–0.71)
0.69 (0.57–0.81) 0.49 (0.31–0.71)
Antiplatelet agentsa
614
0.31 (0.28–0.35)
0.49 (0.45–0.53)
0.57 (0.53–0.61)
0.61 (0.57–0.64)
Antithromboticsb
706
0.32 (0.28–0.35)
0.51 (0.47–0.55)
0.60 (0.57–0.64)
0.65 (0.61–0.68)
Clopidogrel Dipyridamole
PPIs
276
0.67 (0.61–0.72)
0.92 (0.89–0.95)
0.95 (0.92–0.97)
0.97 (0.95–0.99)
H2RAs
103
0.17 (0.11–0.26)
0.25 (0.18–0.35)
0.34 (0.26–0.44)
0.36 (0.28–0.46)
SSRIs
97
0.51 (0.42–0.62)
0.86 (0.78–0.92)
0.91 (0.85–0.96)
0.91 (0.85–0.96)
Steroids
70
0.31 (0.21–0.44)
0.58 (0.46–0.70)
0.68 (0.56–0.79)
0.73 (0.62–0.84)
NSAIDs
347
0.09 (0.06–0.13)
0.15 (0.12–0.20)
0.20 (0.16–0.24)
0.24 (0.20–0.29)
tNSAIDs
295
0.08 (0.06–0.12)
0.12 (0.09–0.17)
0.15 (0.11–0.20)
0.18 (0.14–0.24)
56
0.02 (0.00–0.13)
0.06 (0.02–0.17)
0.11 (0.05–0.23)
0.13 (0.07–0.26)
Coxibs
ASA acetylsalicylic acid, CI confidence interval, coxib cyclooxygenase-2 selective inhibitor, H2RA histamine H2-receptor antagonist, NSAID nonsteroidal anti-inflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor, tNSAID traditional NSAID a
Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole
b
Antithrombotics comprised antiplatelet agents and warfarin
M. E. Sa´ez et al.
a
100
98 97
97
Male Female
89
Re-prescription rate (%)
Fig. 1 Re-prescription rates of study drugs 365 days after upper gastrointestinal bleeding, stratified by (a) sex and (b) age. *Antiplatelet agents comprised acetylsalicylic acid, clopidogrel, and dipyridamole. Antithrombotics comprised antiplatelet agents and warfarin. ASA acetylsalicylic acid, coxib cyclooxygenase-2 selective inhibitor, H2RA histamine H2-receptor antagonist, N/A data not available, NSAID nonsteroidal anti-inflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor, tNSAID traditional NSAID
80
80 73
72 67
65
62
64
60 61
60
66
64
45
42
40
40
32 26
24 24 20
20
19
17 7
0 A
AS
b
n
ari
rf Wa
rel
ole
og
pid
Clo
am
rid
y Dip
tics
ets
tel
pla
ti An
*
†
Is
PP
bo
om
r tith
s
RA H2
RIs
SS
s
ds
AID
roi
Ste
NS
s
98 98
97
94
40–54 years 55–69 years 70–85 years
90 86
Re-prescription rate (%)
85
76
74
72 66
65
60
71
65
68
64
62
Co
An
100
80
s
xib
AID
tNS
63 58
58 50 44
50
47
42
39 39
40
28
25
26
21
20
16
N/A
0 A AS
Wa
N/A
l
n
ri rfa
re og
id
p Clo
da
ri ipy
D
N/A
cs oti
s
le mo
la
p nti
*A
et tel †
b
om thr
Is PP
RA 2
S
ti
An
14
N/A
Is SR
s
H
19
18
s AID
s
St
id ero
NS
s AID
tNS
s
xib
Co
Table 2 Antiplatelet indication and re-prescription rates 365 days after upper gastrointestinal bleeding Indication
Patients (n)
Re-prescription rate (95 % CI)
HR (95 % CI)a
Primary prevention
169 (27.5 %)
0.53 (0.46–0.61)
1.00
Peripheral vascular disease
93 (15.2 %)
0.49 (0.39–0.60)
0.92 (0.64–1.32)
Ischemic heart disease
145 (23.6 %)
0.61 (0.53–0.69)
1.25 (0.93–1.69)
Cerebrovascular disease
106 (17.3 %)
0.67 (0.58–0.76)
1.49 (1.08–2.05)
Unstable angina
12 (1.9 %)
0.79 (0.52–0.96)
2.63 (1.32–5.24)
Myocardial infarction
89 (14.5 %)
0.74 (0.65–0.83)
1.81 (1.31–2.52)
CI confidence interval, HR hazard ratio a Estimates obtained using a Cox proportional hazards model that included age, sex, calendar year, and antiplatelet indication
although the absolute number of ASA re-prescribing was noticeably lower than for all antiplatelets (ESM 3). 3.2 Initiation of Low-dose Acetylsalicylic Acid in the First Year after Upper Gastrointestinal Bleeding We found that 50 of the 1,306 patients with UGIB who had not used ASA in the previous year initiated ASA treatment
during the first year after the UGIB, corresponding to a prescribing rate of 4.0 % (95 % CI 3.1–5.3). During this period after the UGIB, one or more PPI prescriptions were issued in 92 % of these ASA initiators. We found that 99 of the comparison cohort (1,306 date-, age-, and sex-matched controls, who had not used ASA in the previous year) initiated ASA after 1 year of follow-up, corresponding to a prescribing rate of 7.5 % (95 % CI 6.2–9.1). During this period, one or more PPI prescriptions
Patterns in the Use of Low-Dose Acetylsalicylic Acid
were issued in 20 % of these ASA initiators. Overall, initiating ASA treatment was less frequent during the first year after a UGIB (HR 0.53 [95 % CI 0.38–0.75]) than in the comparison cohort without UGIB. The estimate was similar after adjustment for prior comorbidity (HR 0.46 [95 % CI 0.33–0.65]). 3.3 Gastroprotective Agents The cumulative re-prescription rates of gastroprotective agents 1 year after UGIB for current users of H2RAs was 36 %, and for PPIs it was 97 % (Table 1). The lower rate of H2RA re-prescription may be explained by switching to PPIs: 94 % (63/67) of current users of H2RAs who were not re-prescribed H2RAs received a PPI. Among chronic users of gastroprotective agents, re-prescription rates were similar to those of current users overall (ESM 1). Cumulative re-prescription rates at 365 days were 43 % and 99 % for H2RAs and PPIs, respectively. Similar re-prescription rates for H2RAs and PPIs were observed in men and women (Fig. 1a). When analyzed by age, PPI re-prescription rates 1 year after UGIB were similar in all age groups (94 %, 98 %, and 98 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). When analyzed by calendar year, the PPI represcription rate after UGIB was 97 % in 2002 and remained at 97 % in 2007 (ESM 2). Re-prescription of H2RAs increased from 43 % in 2000 to 71 % in 2001, but decreased in subsequent years to 20 % in 2007; this decrease was statistically significant (ptrend = 0.011). 3.4 Anti-Inflammatory Drugs Among the anti-inflammatory drugs studied, the highest represcription rate was for steroids, with 73 % of patients being re-prescribed treatment in the 1-year period after UGIB. In contrast, at 1 year after UGIB, the re-prescription rates for NSAIDs were the lowest among all the drugs analyzed, at 24 %, and were lower for coxibs than for tNSAIDs (Table 1). Among current users of tNSAIDs who did not receive a new prescription, 9 % (20/224) switched to coxibs after UGIB. Among chronic users ([1 year of continuous prescription history before the start date), re-prescription rates of coxibs and tNSAIDs were similar, but re-prescription rates for both were greater than for current users. Cumulative re-prescription rates at 365 days for chronic users were as follows: NSAIDs 38 %; tNSAIDs 29 %; and coxibs 31 % (ESM 1). Similar re-prescription rates for NSAIDS overall and tNSAIDs were observed in men and women (Fig. 1a). When analyzed by age, NSAID re-prescription rates 1 year after UGIB were similar in all age groups (re-prescription
rates were 28 %, 21 %, and 25 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). Steroids were more often re-prescribed to older patients; cumulative re-prescription rates were 76 % for patients aged 70–85 years compared with 50 % for patients aged 40–54 years (no statistically significant difference). When analyzed by calendar year, NSAID re-prescription rates after UGIB were low at 19 % in 2000 and remained low at 20 % in 2007 (ESM 2). Re-prescription of steroids fluctuated, peaking at 91 % in 2002, but falling to 50 % in 2007. 3.5 Selective Serotonin Reuptake Inhibitors For current SSRI users, the re-prescription rate 1 year after UGIB was 91 % (Table 1). Among a subset of chronic SSRI users, re-prescription rates were similar to overall current users at 95 % (ESM 1). SSRIs were re-prescribed in 97 % of men and 89 % of women (Fig. 1a). When analyzed by age, SSRI re-prescription rates 1 year after UGIB were similar in all age groups (86 %, 97 %, and 90 % in patients aged 40–54, 55–69, and 70–85 years, respectively) (Fig. 1b). When analyzed by calendar year, SSRI re-prescription rates after UGIB were generally high, at 71 % in 2000 and 80 % in 2007 (ESM 2). 3.6 Concomitant Use of Proton Pump Inhibitors and Antithrombotics In patients who were prescribed ASA before UGIB (n = 572), only 24 % had received a PPI prescription in the previous year. In patients who were prescribed ASA in the year after UGIB (n = 337), 92 % were prescribed a PPI. Patients who were taking both PPIs and clopidogrel at the time of UGIB had a reduced probability of being represcribed clopidogrel after UGIB compared with patients who used only clopidogrel at the time of UGIB (HR 0.20 [95 % CI 0.06–0.70]) (Table 3). In patients who were taking concomitant PPIs with ASA, warfarin, steroids, or NSAIDs at the time of UGIB, there was a trend towards increased re-prescription of the drug after UGIB compared with patients who used only ASA, warfarin, steroids, or NSAIDs, although this was not statistically significant (HR for ASA 1.26 [95 % CI 0.88–1.81]; warfarin 1.16 [95 % CI 0.55–2.43]; steroids 1.39 [95 % CI 0.68–2.81]; NSAIDs 1.24 [95 % CI 0.66–2.34]) (Table 3). We found that, once patients were prescribed a PPI after UGIB, the overall likelihood of being re-prescribed antithrombotics was increased (HR 2.49 [95 % CI 1.92–3.24]), antiplatelet agents overall (HR 2.38 [95 % CI 1.79–3.18]), ASA (HR 1.68 [95 % CI 1.19–2.37]), and SSRIs (HR 2.33 [95 % CI 1.40–3.89]). For NSAIDs (HR 1.74 [95 % CI
M. E. Sa´ez et al. Table 3 Re-prescription rates and hazard ratios, stratified by concomitant use of proton pump inhibitors at the time of upper gastrointestinal bleeding
Drug
Patients (n)
Patients with re-prescription at 365 days (n)
Re-prescription rate (95 % CI)
HR (95 % CI), adjusted for age, sex, and calendar year
437
171
0.41 (0.37–0.46)
1.00
78
37
0.49 (0.39–0.61)
1.26 (0.88–1.81)
Non-PPI users at start date
41
29
0.74 (0.60–0.86)
1.00
PPI users at start date
11
3
0.30 (0.11–0.67)
0.20 (0.06–0.70)
93
57
0.66 (0.55–0.75)
1.00
12
8
0.71 (0.44–0.93)
1.16 (0.55–2.43)
465
264
0.60 (0.55–0.64)
1.00
85
47
0.58 (0.47–0.69)
0.99 (0.72–1.35)
541
333
0.65 (0.61–0.69)
1.00
96
56
0.61 (0.51–0.71)
0.95 (0.71–1.26)
Non-PPI users at start date
63
55
0.92 (0.83–0.97)
1.00
PPI users at start date
22
20
0.91 (0.75–0.98)
1.00 (0.60–1.68)
Non-PPI users at start date
48
29
0.69 (0.55–0.83)
1.00
PPI users at start date
15
11
0.81 (0.56–0.96)
1.39 (0.68–2.81)
ASA Non-PPI users at start date PPI users at start date Clopidogrela
Warfarin Non-PPI users at start date PPI users at start date Antiplatelet agentsb Non-PPI users at start date PPI users at start date ASA acetylsalicylic acid, CI confidence interval, HR hazard ratio, NSAID nonsteroidal antiinflammatory drug, PPI proton pump inhibitor, SSRI selective serotonin reuptake inhibitor a
At the start date, approximately one-third of clopidogrel users were concomitant users of ASA
b
Antiplatelet agents comprised ASA, clopidogrel, and dipyridamole
c
Antithrombotics comprised antiplatelet agents and warfarin
Antithromboticsc Non-PPI users at start date PPI users at start date SSRIs
Steroids
NSAIDs Non-PPI users at start date PPI users at start date
0.96–3.17]), steroids (HR 1.96 [95 % CI 0.89–4.34]), and warfarin (HR 1.79 [95 % CI 0.93–3.43]), the increased represcription did not reach statistical significance.
4 Discussion This study took place between 2000 and 2007 and reports prescribing practices for low-dose ASA and other therapies after UGIB. As expected, the results showed that antithrombotic use fell after UGIB. ASA re-prescription after UGIB increased slightly over the study period, but fewer than half (43 %) of patients receiving ASA at the time of UGIB received a new prescription in the year following UGIB. Likewise, initiating ASA therapy was half as frequent among individuals with a recent episode of UGIB as among individuals free of UGIB. Re-prescription rates were higher among users of other antithrombotics, including clopidogrel and warfarin, than among users of ASA (although based on small numbers of patients). These re-prescription rates are in line with those reported in a recent study in Denmark, which also found a reduction in re-prescription rates of antithrombotics after
272
57
0.22 (0.18–0.28)
1.00
47
12
0.28 (0.17–0.44)
1.24 (0.66–2.34)
UGIB [20]. This has important clinical implications, since a reduction in antithrombotic use after UGIB would result in an increase in CV risk in these patients [21]. Antithrombotic therapy re-prescription rates decreased with increasing age, indicating caution in re-prescribing such agents in older patients who have had UGIB, despite an increased CV risk in these individuals [22]. Among antiinflammatory drugs, the highest re-prescription rate was observed for steroids. NSAIDs had the lowest represcription rate of the drugs studied, which may be a reflection of the perceived low benefit–risk ratio for these drugs. Although re-prescription of ASA was only 43 % at 1 year after UGIB, the re-prescription of antiplatelet agents overall was 61 %, indicating that many individuals continue or switch antiplatelet agents after UGIB. Among the gastroprotective drugs, the re-prescription rate for PPI users was 97 %, whereas it was only 36 % for H2RA users. In patients who were prescribed ASA before UGIB, only one-quarter had received a PPI prescription in the previous year, but in patients re-prescribed ASA in the year after UGIB, nearly all (92 %) were prescribed a PPI. Incidentally, concomitant use of PPI was identical among
Patterns in the Use of Low-Dose Acetylsalicylic Acid
ASA initiators after UGIB. This reflects guidelines for patients receiving antiplatelet agents (such as ASA), which recommend the use of PPIs for patients who have a history of UGIB, or who have multiple risk factors for bleeding, including combined use of antiplatelet agents and NSAIDs [14, 15]. For patients receiving dual antiplatelet therapy, recent guidelines do not prohibit the use of PPIs in appropriate clinical settings, but highlight the potential risks and benefits from the use of PPIs in combination with clopidogrel [8]. Analysis of the re-prescription rates of the drugs studied in patients who used concomitant PPIs at the time of UGIB indicated that, for concomitant use of PPIs with ASA, warfarin, steroids, or NSAIDs, there was an increased likelihood of re-prescription after UGIB compared with patients who used only ASA, warfarin, steroids, or NSAIDs. In patients who were prescribed a PPI after a UGIB event, there was increased re-prescription of antiplatelet agents and antithrombotics overall, and for the individual study drugs ASA, warfarin, and SSRIs. This is in line with results from other studies, which confirm the effectiveness of PPIs in reducing re-bleeding [23–26]. The current study was undertaken using THIN, a database encompassing a large, representative sample of the UK population, which has been validated for use in epidemiological studies [19], therefore giving our results good external validity. Although over-the-counter low-dose ASA is not recorded in this population, cross-sectional studies have shown that use of low-dose ASA for secondary prevention is commonly prescription based in the UK [27, 28], with patients aged over 60 years receiving free prescriptions, reducing the likelihood of over-the-counter medication use. Some studies have reported rates of overthe-counter low-dose ASA use in secondary prevention populations of between 20 and 40 % [27, 28]; however, more recent data from our own manual review of free-text comments in a random sample of 500 individuals in THIN (aged [40 years, attending primary care services, and without any low-dose ASA prescription) indicated that as few as 5 % of individuals had a mention of overthe-counter low-dose ASA use in their clinical records. As with ASA therapy, some NSAIDs and PPIs are available over the counter and, because such use is not recorded by PCPs, this is a potential limitation of the study, although such practices are likely to be infrequent immediately after an episode of UGIB. In addition, patients leaving the hospital are likely to be provided with medication by the hospital for at least a month. Thus re-prescription rates within 30 days may be missing these hospital prescriptions, although, partly for this reason, the 30-day estimate was not the primary endpoint in the current study. ASA is recommended for long-term maintenance treatment in patients who have a history of CV events [5–8]. It
is reassuring that, in this study, patients who initiated antiplatelet therapy following a myocardial infarction or an episode of unstable angina had a greater probability of continuing antiplatelet therapy after UGIB than those initiating antiplatelet therapy for primary prevention. However, it is clear that not all patients receiving ASA for secondary prevention (who should be receiving continuous long-term ASA) are being re-prescribed antiplatelet agents after UGIB. Discontinuation of ASA has been shown to increase the probability of having a CV event, particularly in the first few days and weeks after stopping the medication [29–31]. We have previously shown a 43 % increase in the risk of CV death or nonfatal myocardial infarction and a 40 % increase in the risk of ischemic stroke or transient ischemic attack among recent low-dose ASA discontinuers [29, 32]. Patients who are not re-prescribed therapy may be left at high risk of secondary CV events if they are not treated with an alternative antithrombotic medication. Several recent studies have indicated that some patients may discontinue their ASA owing to GI adverse effects [33–35]. Assessment of both GI and CV risks versus the benefits of low-dose ASA for individual patients can sometimes be difficult in clinical practice. A recent paper by Lanas et al. [36] reports the development of a risk calculator based on the incidence of CV events and upper GI complications according to the presence of different risk factors. This may prove to be a useful tool to help physicians estimate CV and GI risks and to facilitate the clinical decisionmaking process. Acknowledgments The study was funded in part by AstraZeneca R&D, Mo¨lndal, Sweden. Writing support was provided by Dr Rowan Pearce of Oxford PharmaGenesisTM Ltd and funded by AstraZeneca R&D, Mo¨lndal, Sweden. Conflict of interest statement Marı´a E. Sa´ez, Antonio Gonza´lezPe´rez, and Luis A. Garcı´a Rodrı´guez are employees of the Spanish Centre for Pharmacoepidemiologic Research (CEIFE), which has received research funding from AstraZeneca R&D, Mo¨lndal, Sweden, and Bayer Pharma AG, Berlin, Germany. Luis A. Garcı´a Rodrı´guez has also served as a speaker and an advisory board member for AstraZeneca R&D, Mo¨lndal, Sweden, and Bayer Pharma AG, Berlin, Germany. Saga Johansson and Pe´ter Nagy are employees of AstraZeneca R&D, Mo¨lndal, Sweden.
References 1. Antiplatelet Trialists’ Collaboration. Collaborative overview of randomised trials of antiplatelet therapy—I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ. 1994;308:81–106. 2. Antithrombotic Trialists’ Collaboration. Collaborative metaanalysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ. 2002;324:71–86.
M. E. Sa´ez et al. 3. Patrono C, Coller B, Dalen JE, et al. Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest. 2001;119:39S–63S. 4. Patrono C, Coller B, Dalen JE, et al. Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest. 1998;114:470S–88S. 5. Perk J, De Backer G, Gohlke H, et al. European guidelines on cardiovascular disease prevention in clinical practice (version 2012): The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Eur Heart J. 2012;33:1635–701. 6. Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC). Eur Heart J. 2012;33: 2569–619. 7. O’Gara PT, Kushner FG, Ascheim DD, et al. ACCF/AHA Guideline for the Management of ST-Elevation Myocardial Infarction: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2013;127:e362–425. 8. Anderson JL, Adams CD, Antman EM, et al. 2012 ACCF/AHA focused update incorporated into the ACCF/AHA 2007 guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2013;127:e663–828. 9. Hernandez-Diaz S, Garcia Rodriguez LA. Cardioprotective aspirin users and their excess risk of upper gastrointestinal complications. BMC Med. 2006;4:22. 10. Garcı´a Rodrı´guez LA, Lin KJ, Herna´ndez-Dı´az S, et al. Risk of upper gastrointestinal bleeding with low-dose acetylsalicylic acid alone and in combination with clopidogrel and other medications. Circulation. 2011;123:1108–15. 11. Lanas A, Bajador E, Serrano P, et al. Nitrovasodilators, low-dose aspirin, other nonsteroidal antiinflammatory drugs, and the risk of upper gastrointestinal bleeding. N Engl J Med. 2000;343:834–9. 12. Lanas A, Garcı´a Rodrı´guez LA, Arroyo MT, et al. Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal antiinflammatory drugs, aspirin and combinations. Gut. 2006;55:1731–8. 13. Lanas A. Gastrointestinal bleeding associated with low-dose aspirin use: relevance and management in clinical practice. Expert Opin Drug Saf. 2011;10:45–54. 14. Bhatt DL, Scheiman J, Abraham NS, et al. ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use: a report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents. Circulation. 2008;118: 1894–909. 15. Abraham NS, Hlatky MA, Antman EM, et al. ACCF/ACG/AHA 2010 expert consensus document on the concomitant use of proton pump inhibitors and thienopyridines: a focused update of the ACCF/ACG/AHA 2008 expert consensus document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use. Circulation. 2010;122:2619–33. 16. Garcı´a Rodrı´guez LA, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population. Gastroenterology. 2007; 132:498–506. 17. Herna´ndez-Dı´az S, Garcı´a Rodrı´guez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
tract bleeding/perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med. 2000;160:2093–9. Margulis AV, Garcı´a Rodrı´guez LA, Herna´ndez-Dı´az S. Positive predictive value of computerized medical records for uncomplicated and complicated upper gastrointestinal ulcer. Pharmacoepidemiol Drug Saf. 2009;18:900–9. Lewis JD, Schinnar R, Bilker WB, et al. Validation studies of the health improvement network (THIN) database for pharmacoepidemiology research. Pharmacoepidemiol Drug Saf. 2007;16:393–401. Dall M, Depont Christensen R, de Schaffalitzky Muckadell OB, et al. Re-prescribing of causative drugs in persons discharged after serious drug-induced upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2012;35:948–54. Sung JJ, Lau JY, Ching JY, et al. Continuation of low-dose aspirin therapy in peptic ulcer bleeding: a randomized trial. Ann Intern Med. 2010;152:1–9. Segal WN, Cello JP. Hemorrhage in the upper gastrointestinal tract in the older patient. Am J Gastroenterol. 1997;92:42–6. Lau JY, Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med. 2000;343:310–6. Sung JJ, Barkun A, Kuipers EJ, et al. Intravenous esomeprazole for prevention of recurrent peptic ulcer bleeding: a randomized trial. Ann Intern Med. 2009;150:455–64. Chan FK, Ching JY, Hung LC, et al. Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med. 2005;352:238–44. Masso´ Gonza´lez EL, Garcı´a Rodrı´guez LA. Proton pump inhibitors reduce the long-term risk of recurrent upper gastrointestinal bleeding: an observational study. Aliment Pharmacol Ther. 2008;28:629–37. Bedson J, Whitehurst T, Lewis M, et al. Factors affecting overthe-counter use of aspirin in the secondary prophylaxis of cardiovascular disease. Br J Gen Pract. 2001;51:1001–3. Hopper S, Pierce M. Aspirin after myocardial infarction: the importance of over-the-counter use. Fam Pract. 1998;15(Suppl 1):S10–3. Garcı´a Rodrı´guez LA, Cea-Soriano L, Martı´n-Merino E, et al. Discontinuation of low dose aspirin and risk of myocardial infarction: case–control study in UK primary care. BMJ. 2011;343:d4094. Ferrari E, Benhamou M, Cerboni P, et al. Coronary syndromes following aspirin withdrawal: a special risk for late stent thrombosis. J Am Coll Cardiol. 2005;45:456–9. Biondi-Zoccai GG, Lotrionte M, Agostoni P, et al. A systematic review and meta-analysis on the hazards of discontinuing or not adhering to aspirin among 50,279 patients at risk for coronary artery disease. Eur Heart J. 2006;27:2667–74. Garcı´a Rodrı´guez LA, Cea Soriano L, Hill C, et al. Increased risk of stroke after discontinuation of acetylsalicylic acid: a UK primary care study. Neurology. 2011;76:740–6. Martı´n-Merino E, Johansson S, Bueno H, et al. Discontinuation of low-dose acetylsalicylic acid therapy in UK primary care: incidence and predictors in patients with cardiovascular disease. Prag Obs Res. 2012;3:1–9. Pratt S, Thompson VJ, Elkin EP, et al. The impact of upper gastrointestinal symptoms on nonadherence to, and discontinuation of, low-dose acetylsalicylic acid in patients with cardiovascular risk. Am J Cardiovasc Drugs. 2010;10:281–8. Moberg C, Naesdal J, Svedberg L-E, et al. Impact of gastrointestinal problems on adherence to low-dose acetylsalicylic acid: a quantitative study in patients with cardiovascular risk. Patient. 2011;4:103–13. Lanas A, Polo-Tomas M, Casado-Arroyo R. The aspirin cardiovascular/gastrointestinal risk calculator—a tool to aid clinicians in practice. Aliment Pharmacol Ther. 2013;37:738–48.
