J Gastroenterol (2014) 49:992–1000 DOI 10.1007/s00535-013-0905-z

REVIEW

Aspirin and non-aspirin NSAIDs increase risk of colonic diverticular bleeding: a systematic review and meta-analysis Hiroki Yuhara • Douglas A. Corley • Fumio Nakahara Takayuki Nakajima • Jun Koike • Muneki Igarashi • Takayoshi Suauki • Tetsuya Mine



Received: 9 September 2013 / Accepted: 20 October 2013 / Published online: 14 November 2013 Ó Springer Japan 2013

Abstract Lower gastrointestinal bleeding is a frequent cause of hospitalization, particularly in the elderly, and its incidence appears to be on the rise. Colonic diverticular bleeding is the most common form of lower gastrointestinal bleeding and is responsible for 30–40 % of bleeding episodes. Risk factors associated with diverticular bleeding include obesity, hypertension, anticoagulants, diabetes mellitus, and ischemic heart disease. Recent studies have suggested a relationship between usage of non-steroidal anti-inflammatory drugs (NSAIDs) and colonic diverticular bleeding; however, most studies were small with wide confidence intervals. We identified studies by searching the PubMed and Scopus databases (from inception through 31 December 2012) and by searching bibliographies of relevant articles. Summary relative risks (RRs) with 95 % confidence intervals (CIs) were calculated with fixedeffects and random-effects models. A total of six studies (five case–control studies and one cohort study) met inclusion criteria for analysis. Non-aspirin NSAIDs (NANSAIDs) and aspirin were associated with an increased risk of colonic diverticular bleeding (summary RR = 2.48, 95 % CI 1.86–3.31), with moderate heterogeneity among these studies (P heterogeneity = 0.11, I2 = 44.4 %). Stratification to evaluate the heterogeneity H. Yuhara (&)  F. Nakahara  T. Nakajima  J. Koike  M. Igarashi  T. Suauki  T. Mine Division of Gastroenterology and Hepatology, Department of Internal Medicine, Tokai University School of Medicine, 143 Shimokasuya, Isehara 259-1193, Japan e-mail: [email protected] T. Mine e-mail: [email protected] D. A. Corley Division of Research, Kaiser Permanente, Oakland, USA

123

found that both NANSAIDs (summary RR = 2.24, 95 % CI 1.63–3.09; 5 studies) and aspirin (summary RR = 1.73; 95 % CI 1.31–2.30; 3 studies) were associated with the risk of diverticular bleeding. Aspirin/NANSAIDs use was strongly and consistently associated with an increased risk of colonic diverticular bleeding. Further studies are needed to stratify individuals at risk of diverticular bleeding associated with the use of these agents. Keywords Diverticular bleeding  Lower gastrointestinal bleeding  NSAIDs  Aspirin

Background and aims Acute gastrointestinal bleeding is a common indication for hospital admission. Lower gastrointestinal bleeding, defined as acute or chronic blood loss from a source distal to the ligament of Treitz, accounts for about 15–20 % of gastrointestinal bleeding [1–4]. The incidence of lower gastrointestinal bleeding may be rising, concomitant with the aging of the population and the use of non-steroidal anti-inflammatory drugs (NSAIDs) such as aspirin and the use of anticoagulants [5]. Lower gastrointestinal bleeding is associated with significant mortality, healthcare costs and increased length of hospitalization, emphasizing the need for effective evaluation and treatment [5–8]. Diverticular bleeding is the most common form of lower gastrointestinal bleeding, responsible for 30–40 % of such bleeding episodes; such bleeding is difficult to control endoscopically and is associated with substantial morbidity, the need for invasive procedures, surgery and even mortality [9]. Diverticula represent herniations of the colonic wall at sites where the vasa recta penetrates the colonic wall [10]. The vasa recta are subject to sheer forces

J Gastroenterol (2014) 49:992–1000

where they are draped over the diverticulum. This makes them susceptible to disruption and bleeding. Because the right colon is thin-walled, diverticular bleeding tends to occur more frequently in the ascending colon [11]. Risk factors associated with diverticular bleeding include obesity, hypertension, anticoagulants, diabetes mellitus, and ischemic heart disease [12, 13]. Non-steroidal anti-inflammatory drugs (NSAIDs), including low-dose aspirin, are the most important cause of nonvariceal upper gastrointestinal bleeding in many developed countries. Less well known is the fact that NSAIDs can also induce lower gastrointestinal complications such as bleeding, perforation, stricture, and proteinlosing enteropathy. Studies have suggested a relationship between the usage of NSAIDs, including aspirin, and lower gastrointestinal bleeding, specifically diverticular bleeding [14, 15]. The studies to date, however, had relatively few cases, producing very wide confidence intervals (CIs) and uncertainty regarding the actual magnitude of any association and a limited ability to distinguish any potential risk differences between aspirin and non-aspirin NSAIDS; in addition, the largest case control study suggested there was no excess risk [16]. We conducted a meta-analysis of case–control and cohort studies to assess the association between NSAIDs and colonic diverticular bleeding. We also evaluated whether the association varied by study design, geographic area or type of NSAID (aspirin and non-aspirin NSAIDs).

Methods Search strategy We identified studies using a structured literature search of the PubMed and Scopus databases (from inception through 31 December 2012) for English-language studies showing the following medical subject heading terms or text words: ‘‘diverticular bleeding’’; ‘‘diverticular hemorrhage’’; ‘‘NSAIDs’’; ‘‘non-steroidal anti-inflammatory drugs’’; and ‘‘aspirin’’;. We also reviewed the reference lists of the identified publications and relevant review articles for additional pertinent studies.

993

P value]. Studies were excluded if they provided only an estimate of effect with no means to calculate a confidence interval. Studies were also excluded if the estimates were not adjusted by age and sex. Data extracted included the year of publication, country of the population studied, study design, number of cases, number of exposed and unexposed subjects (cohort studies), number and source of controls (case–control studies), follow-up period (for cohort studies), RR estimates with corresponding CIs, and variables controlled for by matching or in multivariate models. Data abstractions were conducted separately by three authors (H.Y, F.N, and T.N.); discordant results were resolved by consensus. The methodologic quality of case–control and cohort studies was assessed by 2 authors independently (J.K. and M.I.) using the Newcastle–Ottawa scale [17]. In this scale, observational studies were scored across 3 categories: selection (4 questions) and comparability (2 questions) of study groups, and ascertainment of the outcome of interest (3 questions), with all questions with a score of one, except for comparability of study groups, in which separate points were awarded for controlling for age and sex (maximum, 2 points). Any discrepancies were addressed by a joint reevaluation of the original article. Statistical analysis Summary RR estimates were calculated using both the fixed-effects inverse variance weighting method [18] and the random-effects method [19]. We assessed heterogeneity using 2 methods: Cochran’s Q statistic, in which a P value of 0.10 or less was considered evidence of statistically significant heterogeneity, and the I2 statistic, in which a value more than 50 % was suggestive of statistically significant heterogeneity between studies [20]. Publication bias was assessed by constructing a funnel plot and by performing subgroup analyses using Egger and Begg tests [21, 22]. All statistical analyses were undertaken using STATA, version 11.0 (Stata Corp, College Station, TX, USA). Values of P \ 0.05 were considered statistically significant. All statistical tests were two-sided. The analyses and reporting were consistent with the guidelines recommended for meta-analyses of observational studies [23].

Inclusion and exclusion criteria We included published reports that met the following criteria: (1) evaluated risk of first diverticular bleeding; (2) applied a case–control or cohort design; and (3) reported an estimate of relative risk (RR), either an odds ratio (for case–control studies) or a rate or risk ratio (for cohort studies), with a corresponding measure of uncertainty [i.e., 95 % confidence interval (CI), standard error, variance, or

Results Study characteristics Detailed search steps are described in Fig. 1. The electronic search identified 232 abstracts for review; from these, 22 articles potentially met the inclusion criteria and

123

994 Fig. 1 Flowchart of the metaanalysis

J Gastroenterol (2014) 49:992–1000

232 potentially relevant studies identified: citations or abstracts screened for retrieval 214 studies excluded after title or abstract review 18 full studies retrieved for more detailed review 4 additional studies included after reviewing bibliographies of relevant articles

22 potentially appropriate studies assessed for eligibility after full article review

16 studies excluded from analysis 13 studies not reporting bleeding outcomes 2 studies not adjusted by age and sex 1 study from the same study population

6 studies included in the meta-analysis

were retrieved for detailed examination. We excluded two candidate studies [24, 25] that reported RR but were not adjusted by age and sex. We also excluded one study for the same study population [26]. Six articles ultimately met our predefined inclusion criteria and were included in the final analyses [16, 27–31]. Details on these studies are shown in Table 1. The overall methodologic quality of these observational studies was moderate to high. Table 1 depict the methodologic quality of all studies. The j-coefficient of agreement between the reviewers was 1. Most studies use hospital records for exposure ascertainment; outcome assessment was based on record linkage. The temporal relationship between use of NSAIDs and aspirin and the development of diverticular bleeding was established by excluding patients with known diverticular bleeding before nonaspirin NSAIDs (NANSAIDs) and aspirin prescription; only incident cases of diverticular bleeding were included. The included studies variably adjusted for known risk factors for diverticular bleeding including hypertension, diabetes mellitus, obesity, ischemic heart disease, and anticoagulants. Summary results Overall results of the meta-analysis are shown in Fig. 2 and Table 2. All individual studies showed RR estimates above 1.0, and all but one was statistically significant. The

123

summary RR for diverticular bleeding comparing subjects with and without usage of NSAIDs was 2.48 (95 % CI 1.86–3.31) for the 6 studies combined. There was moderate evidence of heterogeneity (Q = 9.00; P = 0.11; I2 = 44.4 %). Sensitivity analysis identified the study by Strate et al. [27] as contributing most to this heterogeneity. In an analysis excluding this study, the summary RR between NSAIDs and diverticular bleeding was similar (summary RR = 3.25; 95 % CI 2.10–5.02) and the test for heterogeneity was not significant (Q = 6.41; P = 0.17; I2 = 37.6 %). Evaluation of heterogeneity We used subgroup analyses to evaluate further several potential sources of heterogeneity. Table 2 shows the results of subgroup meta-analysis by type of NSAIDs, study design, geographical areas, and adjustment for hypertension; these provided homogeneous results for the summaries of aspirin, cohort studies, and patient populations in non-Asian countries; the remaining strata retained moderate heterogeneity (Table 2). The association between NSAIDs and diverticular bleeding was similar for NANSAIDs (summary RR = 2.24; 95 % CI 1.63–3.09; 5 studies) and for aspirin (summary RR = 1.73; 95 % CI 1.31–2.30; 3 studies). The association was stronger in case–control studies (summary RR = 3.25; 95 % CI 2.10–5.02; 5 studies) than in cohort study (summary

No. of cases (control)

RR (95 % CI)

51 (102)

103 (103)

62 (124)

Tsuruoka [28], Japan

Suzuki [16], Japan

Okamoto [27], Japan

NSAIDs 3.4 (1.9–6.2)

NANSAIDs 3.73 (1.26–11.6)

51,529 male dentists, veterinarians, pharmacists, optometrists, osteopathic physicians, and podiatrists, 40–75 years of age (256) Aspirin 1.70 (1.21–2.39)

NANSAIDs 1.74 (1.15–2.64)

NSAIDs 2.02 (1.38–2.96)

Age, sex, study period, BMI, dietary fat, fiber, red meat, nuts, corn, total caloric intake, physical activity

Rectal bleeding attributed to diverticulosis that required hospitalization, blood transfusions, intravenous fluids, surgery, angiography, tagged red blood cell scanning, or endoscopic hemostasis.

Definition of diverticular bleeding

Active bleeding from a diverticulum was observed or when blood clots were found in the colon with diverticula, but no blood was in the terminal ileum and no other demonstrable cause of bleeding was present Active bleeding from a diverticulum was observed or when blood clots were found in the colon with diverticula, but no blood was in the terminal ileum and any other demonstrable cause of bleeding were not present.

Observation of active bleeding from a colonic diverticulum, or presence of blood clots in the colon and colonic diverticulum; absence of blood in the terminal ileum; and no other demonstrable cause of bleeding

Observation of active bleeding from a colonic diverticulum or presence of blood clots over the colon with the presence of colonic diverticulum; absence of blood in the terminal ileum; and no other demonstrable cause of bleeding

Hospital records

Definition of diverticular bleeding

7

NOS score

7

7

7

7

5

NOS score

BMI body mass index, CI confidence interval, NANSAIDs non-aspirin non-steroidal anti-inflammatory drugs, NSAIDs non-steroidal anti-inflammatory drugs, NOS Newcastle–Ottawa Scale, RR relative risk

Strate [29] USA

Cohort studies of NSAIDs and diverticular bleeding incidence based on incidence rate ratios

Controlled variables

Age, sex, distribution of diverticula, HTN, hyperlipidemia, diabetes mellitus, ischemic heart disease or cerebrovascular disease, anticoagulant drugs

Age, sex, colonic diverticulosis, BMI, HTN, ischemic heart disease, hyperlipidemia, warfarin, antihypertensives

Age, sex, ischemic heart disease, hyperlipidemia

Age, sex, location of diverticulosis, aspirin and/or other anticoagulants, HTN

Age, sex, race

Controlled variables

RR (95 % CI)

Aspirin 1.372 (0.52–3.617)

NANSAIDs 9.866 (2.048–47.535)

NANSAIDs 15.6 (1.1–214)

NANSAIDs 2.4 (1.3–4.5)

Aspirin 2.0 (1.1–3.6)

Study population (no. of cases)

44 (88)

Yamada [30], Japan

References, country

53 (N/A)

Wilcox [31], USA

Case–control studies of NSAIDs and diverticular bleeding incidence

References, country

Table 1 Characteristics of studies of NSAIDs and diverticular bleeding

J Gastroenterol (2014) 49:992–1000 995

123

996

J Gastroenterol (2014) 49:992–1000

a

b Study

ES (95% CI)

Wilcox 1997

2.40 (1.29, 4.47)

26.62

Yamada 2006

15.60 (1.12, 217.58)

1.48

Tsuruoka 2010

9.87 (2.05, 47.53)

4.15

Strate 2011

1.74 (1.15, 2.64)

59.43

Okamoto 2012

3.73 (1.23, 11.32)

8.33

Overall (I-squared = 48.6%, p = 0.100)

2.24 (1.63, 3.09)

100.00

.0046

1

Weight (%)

218

c Effect size (95% CI)

Weight (%)

Wilcox 1997

2.0(1.1-3.6)

22.69

Strate 2011

1.70(1.21-2.39)

68.83

Suzuki 2012

1.37(0.52-3.62)

8.48

Overall (I-squared = 0.0%, p = 0.794)

1.73(1.31-2.30)

100.00

Study

.276

1

3.62

Fig. 2 a Association between non-steroidal anti-inflammatory drugs (NSAIDs) and risk of diverticular bleeding incidence in 6 studies. b Association between the use of non-aspirin non-steroidal anti-

inflammatory drugs (NANSAIDs) and risk of diverticular bleeding in 5 studies. c Association between the use of aspirin and risk of diverticular bleeding in 3 studies

RR = 2.02; 95 % CI 1.38–2.96; one study). Subgroup meta-analysis by geographical areas revealed a stronger association in studies from Asian countries (summary

RR = 3.08; 95 % CI 1.62–5.85; 4 studies) than from Western countries (summary RR = 2.35; 95 % CI 1.71–3.24; 2 studies).

123

J Gastroenterol (2014) 49:992–1000

997

Table 2 Summary relative risk (RR) estimates and 95 % CIs for case–control and cohort studies of associations between NSAIDs and diverticular bleeding incidence by type of NSAID, study design, geographical area, and adjustment for hypertension Subgroup

No. of studies

Summary RR (95 % CI) Fixed effect model

Q

P heterogeneity

I2 statistic (%)

Summary RR (95 % CI) Random effect model

All studies

6

2.48 (1.86–3.31)

9.00

0.11

44.4

2.85 (1.76–4.61)

NANSAIDs

5

2.24 (1.63–3.09)

7.78

0.10

48.6

2.87 (1.62–5.07)

Aspirin

3

1.73 (1.31–2.30)

0.46

0.79

0.0

1.73 (1.31–2.30)

Case–control

5

3.25 (2.10–5.02)

6.41

0.17

37.6

3.44 (1.83–6.48)

Cohort

1

2.02 (1.38–2.96)

0.0

Asian

4

3.08 (1.62–5.85)

6.36

0.10

52.8

3.89 (1.40–10.80)

Western

2

2.35 (1.71–3.24)

2.10

0.15

52.4

2.49 (1.51–4.11)

Type of NSAIDs

Study design 2.02 (1.38–2.96)

Geographical area

.5 1.5

1

s.e. of lnrr

0

HTN hypertension, NANSAIDs non-aspirin non-steroidal anti-inflammatory drugs, NSAIDs non-steroidal anti-inflammatory drugs

-2

0

2

4

lnrr

Fig. 3 Funnel plot with pseudo 95 % confidence limits

Publication bias Visual inspection of the Begg funnel plot for diverticular bleeding did not demonstrate the asymmetry typically associated with publication bias (Fig. 3). Evidence of publication bias was also not seen with the Egger or Begg tests (Egger P = 0.147).

Discussion This meta-analysis shows a significant association between the risk of diverticular hemorrhage and prior use of NSAIDs. The association between the use of NSAIDs and the risk of diverticular bleeding was statistically significant for NANSAIDs (summary RR = 2.24; 95 % CI 1.63–3.09;

5 studies) and for aspirin (summary RR = 1.73; 95 % CI 1.31–2.30; 3 studies). Statistical heterogeneity was seen in several of our analyses, and much of this is likely due to differences in study design, study population, and definition of NSAIDs. Statistical heterogeneity was low in our analyses of studies which evaluated aspirin and was lower in cohort studies than in case control studies. This suggests that a potential difference in bleeding risk between NANSAIDs and aspirin in some studies and some amplification of risk associated with the case control design accounted for much of the heterogeneity observed; this latter finding is not unexpected as the odds ratio calculated in case–control studies may overestimate risk compared with a risk ratio. Data from several sources suggest that the association between NSAIDs including aspirin and diverticular bleeding is biologically plausible, although the exact mechanisms underlying diverticular bleeding are not well understood. Aspirin and NANSAIDs inhibit platelet cyclooxygenase, thereby blocking the formation of thromboxane A2 [32]. These drugs produce a tendency toward systemic bleeding by impairing thromboxane-dependent platelet aggregation and consequently prolonging bleeding time. Diverticular bleeding occurs when an arteriole ruptures into the colon lumen, and commonly involves local ulceration in the absence of inflammation [33]. Several limitations of this study must be considered. First, all of the studies that have assessed the association between NSAIDs including aspirin and diverticular bleeding have been observational in design, which are susceptible to bias and confounding, especially as compared with randomized trials [34]. However, several factors suggest that, despite the inherent limitations of

123

998

observational studies, our finding that both aspirin and NANSAIDs are associated with an increased risk of diverticular bleeding may represent a real association. These include a relatively large effect size and fairly homogeneous results among studies that adjusted for important confounders. Second, the studies may have misclassified some patients regarding the source of bleeding. Bleeding is often attributed to the presence of diverticula when no other source is found and the clinical history is consistent with a diverticular source; however, there is no clear reason why this misclassification would be differential between NSAID users vs. nonusers. Spontaneous resolution of bleeding from colonic diverticula is common, with overall rates of 76 % [35]. Jensen et al. at the Center for Ulcer Research and Education (CURE) Hemostasis Research Unit reported that, among patients hospitalized with colonic diverticulosis and severe hematochezia who were studied prospectively with urgent colonoscopy after purge, approximately 50 % were bleeding from non-diverticular sources, about 30 % had presumed diverticular hemorrhage, and 20 % had evidence of definitive diverticular hemorrhage (such as active bleeding, a non-bleeding visible vessel or an adherent clot) [36, 37]. In addition, balloon enteroscopy and capsule endoscopy studies have shown an increased risk of small bowel injury among NSAIDs users [38, 39]. Patients in the current studies had procedures to exclude other likely colonic sources of bleeding, but we cannot exclude that for some patients small intestinal bleeding was misdiagnosed as presumptive diverticular hemorrhage. Third, most studies reviewed did not evaluate the dose, frequency or duration of NSAID use. Low-dose aspirin, commonly defined as a dose of 75–325 mg daily, is used for primary and secondary prevention of cardiovascular events such as myocardial infarction. In this meta-analysis, four studies evaluated aspirin usage, but only the study by Strate et al. [29] evaluated the risk of diverticular bleeding by dose of aspirin and showed that the association did not display a linear dose relationship. They also showed that an increasing duration of regular aspirin and non-aspirin NSAID use was associated with a greater risk of diverticular bleeding. Fourth, most studies reviewed did not specify the type of NSAIDs (nonselective NSAIDs and cyclooxygenase-2 inhibitors). Systematic review by Laine et al. [40] reported that a lower rate of low GI injury for cyclooxygenase-2 inhibitors as compared with nonselective NSAIDs. In addition, nonselective NSAIDs vary widely in their propensity to cause clinically significant GI events and for many NSAIDs there is also a strong dose response effect with regard to risk. This may distort the true relationship between NSAIDs and risk of diverticular bleeding, if these differences are not taken into account. In addition, most

123

J Gastroenterol (2014) 49:992–1000

studies did not take over-the-counter NSAIDs into account, and, therefore, misclassification might ensure and bias study results towards the null. Fifth, hypertension is potentially the most important confounder of the positive association between NSAIDs and diverticular bleeding risk. Although some of the studies included in this meta-analysis controlled for hypertension, we were unable to control for other potential confounders of the association between NSAIDs and diverticular bleeding including obesity, diabetes mellitus, and ischemic heart disease. If these factors were strongly associated with both diverticular disease and aspirin use, the lack of adjustment for these variables could have contributed to a non-causal association between NSAIDs and diverticular bleeding, although these factors may be partially adjusted for by their associations with hypertension. We did not find that adjustment for hypertension had a substantial impact on the relationship between NSAIDs and diverticular bleeding. Sixth, a limited sample of studies were available for the analysis, which yielded small number of studies in each of the subgroup analyses. This may have contributed to the effect of size variance found within most of the study aggregations. A final limitation is the fact that most studies included in this meta-analysis were conducted in Japan, meaning that their results may not be generalizable to a Western population, due to potential underlying differences in patient characteristics, genetics or environment. In fact, ethnic differences exist in the location of diverticulosis, including differences in the distribution of diverticulosis between Western and Asian populations: Left-sided diverticular disease predominates in Western populations, whereas right-sided disease in the colon is commonly reported in Asians [41]. Some right-sided colon diverticulosis has also been postulated to be congenital, unlike the development of sigmoid diverticular disease, which is thought to be acquired as a result of elevated intraluminal pressure within the colon attributable to inadequate intake of dietary fiber [42, 43]. It is not known whether differences in the distribution or density of diverticula affect the risk of acute or chronic complications of diverticulosis. One study estimated that acute complications occur in 15–20 % of persons with predominantly left-sided diverticular disease in the West [44] and compared with only 1.0–4.5 % of persons with predominantly right-sided diverticular disease in Southeast Asia; rates may be even lower among patients with only one or two diverticula [45]. As differences in etiology may exist between right- and left-sided diverticulosis, differences in risk of bleeding may also exist. In this meta-analysis, we could not perform subgroup analysis based on the race and location of diverticular bleeding. The true relationship between NSAIDs and diverticular

J Gastroenterol (2014) 49:992–1000

bleeding may thus be distorted if these differences are not taken into account. The use of NSAIDs is widespread because of their numerous applications and accessibility as an analgesic and anti-inflammatory drug [46]. Low-dose aspirin is also routinely used in the primary and secondary prophylaxis of cardiovascular disease. Where the elderly population is increasing, the consumption of non-aspirin NSAIDs and low-dose aspirin seems likely to increase in the future. In addition, an expanding body of evidence suggests that NSAIDs, including aspirin, may play a prophylactic role in gastrointestinal oncology [47]. Furthermore, there is no established prophylaxis for lower gastrointestinal bleeding including diverticular bleeding, while proton pump inhibitors and Helicobacter pylori eradication can reduce the risk of upper gastrointestinal bleeding. Future studies are needed to better identify individuals at risk of diverticular bleeding, and to develop strategies to mitigate the risk of lower gastrointestinal bleeding. Conflict of interest of interest.

The authors declare that they have no conflict

References 1. Bass BL, Alvarez C. Acute gastrointestinal hemorrhage. In: Townsend CM, Sabiston DC, editors. Textbook of surgery. Philadelphia: Saunders; 2001. p. 816–35. 2. Longstreth GF. Epidemiology and outcome of patients hospitalized with acute lower gastrointestinal hemorrhage: a populationbased study. Am J Gastroenterol. 1997;92(3):419–24. 3. Vernava AM 3rd, Moore BA, Longo WE, et al. Lower gastrointestinal bleeding. Dis Colon Rectum. 1997;40(7):846–58. 4. Gostout CJ, Wang KK, Ahlquist DA, et al. Acute gastrointestinal bleeding. Experience of a specialized management team. J Clin Gastroenterol. 1992;14:260–7. 5. Lanas A, Garcı´a-Rodrı´guez LA, Polo-Toma´s M, et al. The changing face of hospitalisation due to gastrointestinal bleeding and perforation. Aliment Pharmacol Ther. 2011;33:585–91. 6. Rı´os A, Montoya MJ, Rodrı´guez JM, et al. Acute lower gastrointestinal hemorrhages in geriatric patients. Dig Dis Sci. 2005;50:898–904. 7. Comay D, Marshall JK. Resource utilization for acute lower gastrointestinal hemorrhage: the Ontario GI bleed study. Can J Gastroenterol. 2002;16:677–82. 8. Lhewa DY, Strate LL. Pros and cons of colonoscopy in management of acute lower gastrointestinal bleeding. World J Gastroenterol. 2012;18(11):1185–90. 9. Zuccaro G Jr. Management of the adult patient with acute lower gastrointestinal bleeding. American College of Gastroenterology. Practice Parameters Committee. Am J Gastroenterol. 1998;93(8):1202–8. 10. Hall J, Hammerich K, Roberts P. New paradigms in the management of diverticular disease. Curr Probl Surg. 2010;47(9):680–735. 11. Wong SK, Ho YH, Leong AP, et al. Clinical behavior of complicated right-sided and left-sided diverticulosis. Dis Colon Rectum. 1997;40(3):344–8.

999 12. Strate LL, Liu YL, Aldoori WH, et al. Obesity increases the risks of diverticulitis and diverticular bleeding. Gastroenterology. 2009;136(1):115–22. 13. Parsa F, Gordon HE, Wilson SE. Bleeding diverticulosis of the colon: a review of 83 cases. Dis Colon Rectum. 1975;18:37–41. 14. Laine L, Connors LG, Reicin A et al. Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology 2003;124(2):288e92. 15. Foutch PG. Diverticular bleeding: are nonsteroidal anti-inflammatory drugs risk factors for hemorrhage and can colonoscopy predict outcome for patients? Am J Gastroenterol 1995;90(10):1779e84. 16. Suzuki K, Uchiyama S, Imajyo K, et al. Risk factors for colonic diverticular hemorrhage: Japanese multicenter study. Digestion. 2012;85(4):261–5. 17. Wells GA, Shea B, O’Connell D, Peterson J, et al. The Newcastle–Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. 2012. http://www.ohri.ca/ programs/clinical_epidemiology/oxford.asp. Accessed 5 Sept 2012. 18. Greenland S. Meta-analysis. In: Rothman K, Greenland S, editors. Modern epidemiology. 2nd ed. Philadelphia: Lippincott Raven; 1998. p. 643–73. 19. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–88. 20. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–60. 21. Egger M, Davey Smith G, Schneider M, et al. Bias in metaanalysis detected by a simple, graphical test. Br Med J. 1997;315:629–34. 22. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50: 1088–101. 23. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283(15):2008–12. 24. Jansen A, Harenberg S, Grenda U, et al. Risk factors for colonic diverticular bleeding: a Westernized community based hospital study. World J Gastroenterol. 2009;15(4):457–61. 25. Niikura R, Nagata N, Akiyama J, et al. Hypertension and concomitant arteriosclerotic diseases are risk factors for colonic diverticular bleeding: a case–control study. Int J Colorectal Dis. 2012;27(9):1137–43. 26. Aldoori WH, Giovannucci EL, Rimm EB, et al. Use of acetaminophen and nonsteroidal anti-inflammatory drugs: a prospective study and the risk of symptomatic diverticular disease in men. Arch Fam Med. 1998;7(3):255–60. 27. Okamoto T, Watabe H, Yamada A, et al. The association between arteriosclerosis related diseases and diverticular bleeding. Int J Colorectal Dis. 2012;27(9):1161–6. 28. Tsuruoka N, Iwakiri R, Hara M, et al. NSAIDs are a significant risk factor for colonic diverticular hemorrhage in elder patients: evaluation by a case–control study. J Gastroenterol Hepatol. 2011;26(6):1047–52. 29. Strate LL, Liu YL, Huang ES, et al. Use of aspirin or nonsteroidal anti-inflammatory drugs increases risk for diverticulitis and diverticular bleeding. Gastroenterology. 2011;140(5): 1427–33. 30. Yamada A, Sugimoto T, Kondo S, et al. Assessment of the risk factors for colonic diverticular hemorrhage. Dis Colon Rectum. 2008;51(1):116–20. 31. Wilcox CM, Alexander LN, Cotsonis GA, et al. Nonsteroidal antiinflammatory drugs are associated with both upper and lower gastrointestinal bleeding. Dig Dis Sci. 1997;42(5):990–7.

123

1000 32. Schafer AI. Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis. J Clin Pharmacol. 1995;35(3):209–19. 33. Meyers MA, Alonso DR, Baer JW. Pathogenesis of massively bleeding colonic diverticulosis: new observations. Am J Roentgenol. 1976;127:901–8. 34. Bjarnason I, Hayllar J, MacPherson AJ, et al. Side effects of nonsteroidal anti-inflammatory drugs on the small and large intestine in humans. Gastroenterology. 1993;104:1832–47. 35. Guyatt GH, Oxman AD, Kunz R, GRADE Working Group, et al. What is ‘‘quality of evidence’’ and why is it important to clinicians? Br Med J. 2008;336:995–8. 36. McGuire HH Jr. Bleeding colonic diverticula. A reappraisal of natural history and management. Ann Surg. 1994;220:653–6. 37. Jensen DM, Machicado GA, Jutabha R, et al. Urgent colonoscopy for the diagnosis and treatment of severe diverticular hemorrhage. N Engl J Med. 2000;342(2):78–82. 38. Higuchi K, Umegaki E, Watanabe T, et al. Present status and strategy of NSAIDs-induced small bowel injury. J Gastroenterol. 2009;44(9):879–88. 39. Maiden L. Capsule endoscopic diagnosis of nonsteroidal antiinflammatory drug-induced enteropathy. J Gastroenterol. 2009; 44(Suppl 19):64–71.

123

J Gastroenterol (2014) 49:992–1000 40. Laine L, Smith R, Min K, Chen C, Dubois RW. Systematic review: the lower gastrointestinal adverse effects of non-steroidal anti-inflammatory drugs. Aliment Pharmacol Ther. 2006;24(5): 751–67. 41. Hale WB. NDSG. Colonoscopy in the diagnosis and management of diverticular disease. J Clin Gastroenterol. 2008;42(10):1142–4. 42. Fong SS, Tan EY, Foo A, et al. The changing trend of diverticular disease in a developing nation. Colorectal Dis. 2011;13(3):312–6. 43. Arfwidsson S, Knock NG, Lehmann L, et al. Pathogenesis of multiple diverticula of the sigmoid colon in diverticular disease. Acta Chir Scand Suppl. 1964;342:341–68. 44. Parks TG. Natural history of diverticular disease of the colon. Clin Gastroenterol. 1975;4:53–69. 45. Miura S, Kodaira S, Shatari T, et al. Recent trends in diverticulosis of the right colon in Japan: retrospective review in a regional hospital. Dis Colon Rectum. 2000;43:1383–9. 46. Leung Ki EL, Chan FK. Helicobacter pylori infection and nonsteroidal anti-inflammatory drug use: eradication, acid-reducing therapy, or both? Clin Gastroenterol Hepatol. 2012;10(8):831–6. 47. Thun MJ, Jacobs EJ, Patrono C. The role of aspirin in cancer prevention. Nat Rev Clin Oncol. 2012;9(5):259–67.

Aspirin and non-aspirin NSAIDs increase risk of colonic diverticular bleeding: a systematic review and meta-analysis.

Lower gastrointestinal bleeding is a frequent cause of hospitalization, particularly in the elderly, and its incidence appears to be on the rise. Colo...
386KB Sizes 0 Downloads 0 Views