Aspirin for the Primary Prevention of Cardiovascular Disease and Colorectal Cancer: A Decision Analysis for the U.S. Preventive Services Task Force

Steven P. Dehmer; Michael V. Maciosek; Thomas J. Flottemesch; Amy B. LaFrance; Evelyn P. Whitlock

doi:10.7326/M15-2129

Original Research |21 June 2016

Aspirin for the Primary Prevention of Cardiovascular Disease and Colorectal Cancer: A Decision Analysis for the U.S. Preventive Services Task Force Free

Steven P. Dehmer, PhD; Michael V. Maciosek, PhD; Thomas J. Flottemesch, PhD; Amy B. LaFrance, MPH; Evelyn P. Whitlock, MD, MPH

Article, Author, and Disclosure Information

This article was published at www.annals.org on 12 April 2016.

From HealthPartners Institute, Minneapolis, Minnesota, and Kaiser Permanente Center for Health Research, Portland, Oregon.
Disclaimer: The views expressed in this article do not represent and should not be construed to represent a determination or policy of the Agency for Healthcare Research and Quality (AHRQ) or the U.S. Department of Health and Human Services.
Acknowledgment: The authors gratefully acknowledge the following persons for their contributions to this project: Robert McNellis, PA, MPH, at the AHRQ; Kirsten Bibbins-Domingo, PhD, MD, MAS, Michael L. LeFevre, MD, MSPH, Douglas K. Owens, MD, MS, and Michael P. Pignone, MD, MPH, of the USPSTF; Janelle M. Guirguis-Blake, MD, Jessica Chubak, PhD, MBHL, Melissa L. Anderson, MS, Tracy Beil, MS, Diana S.M. Buist, PhD, MPH, Brittany U. Burda, MPH, Corinne V. Evans, MPP, Alisha Feightner, MPH, Aruna Kamineni, PhD, MPH, Elizabeth A. O'Connor, PhD, Maya G. Rowland, MPH, Caitlyn A. Senger, MPH, and Selvi B. Williams, MD, MPH, with the Kaiser Permanente Research Affiliates Evidence-based Practice Center; and Logan H. Stuck, MS, at the HealthPartners Institute. The following persons provided peer review of the work plan, full evidence report, or both: Dong-Yun Kim, PhD, William Lawrence, MD, MS, Michael Pignone, MD, MPH, Glen Taksler, PhD, and Steven Teutsch, MD, MPH.
Financial Support: By contract HHSA-290-2012-00015-EPC4, Task Order 4, from AHRQ.
Disclosures: The authors report a contract with AHRQ during the conduct of the study. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M15-2129.
Editors' Disclosures: Christine Laine, MD, MPH, Editor in Chief, reports that she has no financial relationships or interests to disclose. Darren B. Taichman, MD, PhD, Executive Deputy Editor, reports that he has no financial relationships or interests to disclose. Cynthia D. Mulrow, MD, MSc, Senior Deputy Editor, reports that she has no relationships or interests to disclose. Deborah Cotton, MD, MPH, Deputy Editor, reports that she has no financial relationships or interest to disclose. Jaya K. Rao, MD, MHS, Deputy Editor, reports that she has stock holdings/options in Eli Lilly and Pfizer. Sankey V. Williams, MD, Deputy Editor, reports that he has no financial relationships or interests to disclose. Catharine B. Stack, PhD, MS, Deputy Editor for Statistics, reports that she has stock holdings in Pfizer and Johnson & Johnson.
Reproducible Research Statement:Study protocol: Available from Dr. Dehmer (e-mail, steven.p.dehmer@healthpartners.com). Statistical code and data set: Please contact Dr. Dehmer (e-mail, steven.p.dehmer@healthpartners.com) for availability (some restrictions may apply).
Requests for Single Reprints: Reprints are available from the AHRQ Web site (www.ahrq.gov).
Current Author Addresses: Drs. Dehmer and Maciosek and Ms. LaFrance: HealthPartners Institute, Mailstop 21111R, PO Box 1524, Minneapolis, MN 55440.
Dr. Flottemesch: Truven Health Analytics, 777 East Eisenhower Parkway, Ann Arbor, MI 48108.
Dr. Whitlock: Patient-Centered Outcomes Research Institute, 1828 L Street, Northwest, Suite 900, Washington, DC 20036.
Author Contributions: Conception and design: S.P. Dehmer, M.V. Maciosek, T.J. Flottemesch, E.P. Whitlock.
Analysis and interpretation of data: S.P. Dehmer, M.V. Maciosek, T.J. Flottemesch, A.B. LaFrance, E.P. Whitlock.
Drafting of the article: S.P. Dehmer, M.V. Maciosek.
Critical revision of the article for important intellectual content: S.P. Dehmer, M.V. Maciosek, T.J. Flottemesch, LaFrance, E.P. Whitlock.
Final approval of the article: S.P. Dehmer, M.V. Maciosek, T.J. Flottemesch, A.B. LaFrance, E.P. Whitlock.
Statistical expertise: T.J. Flottemesch, S.P. Dehmer.
Obtaining of funding: M.V. Maciosek, E.P. Whitlock.
Administrative, technical, or logistic support: A.B. LaFrance.
Collection and assembly of data: S.P. Dehmer, T.J. Flottemesch, E.P. Whitlock.

Abstract

Background:

Evidence indicates that aspirin is effective for the primary prevention of cardiovascular disease (CVD) and colorectal cancer (CRC) but also increases the risk for gastrointestinal (GI) and cerebral hemorrhages.

Objective:

To assess the net balance of benefits and harms from routine aspirin use across clinically relevant age, sex, and CVD risk groups.

Design:

Decision analysis using a microsimulation model.

Data Sources:

3 systematic evidence reviews.

Target Population:

Men and women aged 40 to 79 years with a 10-year CVD risk of 20% or less, and no history of CVD and without elevated risk for GI or cerebral hemorrhages that would contraindicate aspirin use.

Time Horizon:

Lifetime, 20 years, and 10 years.

Perspective:

Clinical.

Intervention:

Low-dose aspirin (≤100 mg/d).

Outcome Measures:

Primary outcomes are length and quality of life measured in net life-years and quality-adjusted life-years. Benefits include reduced nonfatal myocardial infarction, nonfatal ischemic stroke, fatal CVD, CRC incidence, and CRC mortality. Harms include increased fatal and nonfatal GI bleeding and hemorrhagic stroke.

Results of Base-Case Analysis:

Lifetime net quality-adjusted life-years are positive for most adults initiating aspirin at ages 40 to 69 years, and life expectancy gains are expected for most men and women initiating aspirin at ages 40 to 59 years and 60 to 69 years with higher CVD risk. Harms may exceed benefits for persons starting aspirin in their 70s and for many during the first 10 to 20 years of use.

Results of Sensitivity Analysis:

Results are most sensitive to the relative risk for hemorrhagic stroke and CVD mortality but are affected by all relative risk estimates, baseline GI bleeding incidence and case-fatality rates, and disutilities associated with aspirin use.

Limitations:

Aspirin effects by age are uncertain. Stroke benefits are conservatively estimated. Gastrointestinal bleeding incidence and case-fatality rates account only for age and sex.

Conclusion:

Lifetime aspirin use for primary prevention initiated at younger ages (40 to 69 years) and in persons with higher CVD risk shows the greatest potential for positive net benefit.

Primary Funding Source:

Agency for Healthcare Research and Quality.

Editors’ Notes

Context

Benefits and harms of routine aspirin use vary among individuals.

Contribution

This modeling study suggested that lifetime aspirin use for primary prevention of cardiovascular disease (CVD) and colorectal cancer (CRC) had potential net benefits for most men and women who did not have elevated bleeding risk and initiated aspirin use at ages 40 to 69 years. Overall benefits did not outweigh harms for persons in their 70s with a 10-year CVD risk of 20% or less.

Caution

Estimates of aspirin effects by age were uncertain.

Implication

Middle-aged men and women without elevated risk for gastrointestinal or cerebral hemorrhage should consider long-term aspirin use to prevent CVD and CRC.

Evidence for the effectiveness of aspirin in preventing recurrent complications from heart disease and stroke (secondary prevention) is strong (1, 2), but evidence for aspirin's net benefit in preventing cardiovascular disease (CVD) and cancer, including colorectal cancer (CRC), in healthy persons (primary prevention) has been mixed (2–8). Three recent systematic reviews conducted on behalf of the U.S. Preventive Services Task Force (USPSTF) investigated current evidence for the benefits and harms of aspirin for primary prevention of CVD, on all-cause mortality, for all types of cancer, and for CRC (9–14). These reviews reaffirm evidence of aspirin's effectiveness—no longer differing by sex—in preventing first-time myocardial infarction (MI) and ischemic stroke and find new evidence indicating its effectiveness in CRC prevention. However, the updated reviews also reaffirm aspirin's role in increasing the risk for major gastrointestinal (GI) bleeding and hemorrhagic stroke.

The central clinical dilemma in determining the appropriateness of aspirin for the primary prevention of CVD and CRC is an uncertain relationship between the benefits and harms of long-term aspirin use. Therefore, we conducted a decision analysis using simulation modeling to assess the expected net benefit of aspirin use for primary prevention across clinically relevant population groups defined by their age, sex, and underlying CVD risk characteristics. This study was initiated by the USPSTF to support the update (15) of its recommendations on using aspirin for primary prevention (3, 4).

Methods

Model Description

We conducted study analyses using the Health Partners Institute ModelHealth: CVD microsimulation model. This annual-cycle microsimulation model was parameterized to estimate the person-level natural history of cardiovascular risk factors and the lifetime incidence of CVD events in a cross-section representative of the U.S. population. A CRC incidence and case-fatality natural history module was added to our model for this study. A detailed description of the model, implemented by using Visual Basic 6.0 (Microsoft) and Microsoft Excel, can be found in the Supplement.

Target Population

Aspirin for primary prevention was assessed independently for men and women across four 10-year age bands (40 to 49, 50 to 59, 60 to 69, and 70 to 79 years) and baseline 10-year CVD risk bands (ranging from 1% to 20%). Baseline 10-year CVD risk was rounded to the nearest integer and estimated using the American College of Cardiology/American Heart Association risk calculator for the first hard atherosclerotic CVD event (16). The calculation of CVD risk at baseline is independent from the event rates predicted by the model. For each age, sex, and baseline CVD risk band, simulated persons were randomly oversampled from population characteristics representative of the U.S. population. For men aged 60 to 79 years and women aged 70 to 79 years, 10-year low-risk bands that are rarely or never observed in NHANES (National Health and Nutrition Examination Survey) of the U.S. population were excluded.

Initial demographic characteristics were drawn from the U.S. Census (17). Initial body mass index, systolic blood pressure, high- and low-density lipoprotein cholesterol levels, and diabetes status were derived from the 2001 to 2010 NHANES data (18–22). Initial smoking status was derived from the 2007 National Health Interview Survey (23) and calibrated to projections from the Congressional Budget Office (24). All persons were assumed to be free of CVD and CRC and to have nonelevated bleeding risk at baseline (defined by the absence of any factors for which a clinical provider would deem aspirin unsafe, such as history of GI or intracranial bleeding or concurrent use of other medications that increase bleeding risk).

Study Perspective

Analyses were conducted from a clinical perspective with respect to health outcomes associated with aspirin use. Costs were not considered.

Time Horizon

The primary time horizon is over a lifetime, which we defined at the person level as the “age to death or age 100” in order to fully account for ongoing benefits and harms (25). Time horizons of 10 and 20 years are included for their practical relevance.

Choice of Intervention

Findings from the 3 coordinated, companion systematic evidence reviews were integral to the parameter assumptions and model design in this study (9–14). The reviews found evidence that daily aspirin use reduces the risk for nonfatal MI, nonfatal stroke, and 10-year (and greater) CRC incidence and mortality. Aspirin also was found to increase the risk for hemorrhagic stroke and major GI bleeding. The best balance of cardiovascular benefits and harms was reflected in aspirin doses of 100 mg/d or less (low dose). Benefits with respect to CRC incidence were not strongly correlated with dose or prior CVD status, and therefore higher aspirin dose and secondary prevention trials were included in deriving this parameter. No clear evidence was found that aspirin changes the relative risk (RR) for CVD death, fatal GI bleeding, all-cause mortality, or other types of cancer or that aspirin effects differ by age or, in contrast to prior USPSTF findings (4, 26), sex. Evidence reviews also informed baseline levels of GI bleeding risk and selection of the American College of Cardiology/American Heart Association risk calculator to specify baseline CVD risk in the model (16).

Intervention Effects

Effects from using aspirin for primary prevention were modeled as RR modifications to the annual probability of an event. The CVD and bleeding RRs were derived from 8 trials about low-dose aspirin for primary prevention (12, 27–34). The effect of aspirin on the RR for CRC incidence after 10 years of continuous use was estimated from 3 aspirin trials (13, 35, 36) (Table 1). Only a few low-dose aspirin trials independently reported ischemic stroke events (9); therefore, we used a combined stroke measure that included hemorrhagic stroke events to approximate the effect of aspirin on ischemic stroke, resulting in a conservative estimate of ischemic stroke benefits. All non-CRC benefits and harms with aspirin initiation are assumed to take effect immediately, and all RRs are assumed to return to 1.00 with aspirin discontinuation. Indirect effects of aspirin on disease incidence and mortality may arise when the prevention or occurrence of an initial event alters the disease progression probabilities for subsequent events.

Table 1. Key Aspirin Benefit and Harm Parameter Values*

Health utilities for outcomes affected by aspirin use were estimated using literature sources (37–46) (Appendix Table 1). Living without a CVD condition or CRC was given a health utility of 0.872. All other health utility weights were applied multiplicatively to that baseline. Disutilities from MI and GI bleeding events were applied only during the year an event occurred. In the base-case analysis, no disutility was applied to taking aspirin daily, but 2 alternative scenarios with aspirin disutilities were considered in sensitivity analyses.

Appendix Table 1. Health Utility Weights*

Analysis Design and Outcomes

All analyses compared outcomes of a simulated population routinely using aspirin for the primary prevention of CVD and CRC with the same population, all else held equal, not using aspirin for primary prevention (Figure). Primary outcomes are the net difference in undiscounted life-years and quality-adjusted life-years (QALYs), but all modeled benefits and harms were measured. Aspirin was initiated or continued at contemporary rates for secondary prevention in both simulation groups. It was discontinued permanently in both groups after any major GI bleeding or hemorrhagic stroke event. Model simulations were independently conducted with a 100 000-person sample for each age, sex, and baseline CVD risk group.

Figure.

Simulation model and analysis design.

BMI = body mass index; BP = blood pressure; CRC = colorectal cancer; CVD = cardiovascular disease; HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol; SBP = systolic blood pressure.

Baseline Event Rates and Model Validation

Baseline rates of CVD events are generated by the combination of population characteristics at model initiation, the natural progression of CVD risk factors as persons age, and the model's risk equations for disease. Appendix Table 2 compares rates of MI and ischemic stroke generated by the model with corresponding rates observed in NHANES (18–22) for external validation of our model's natural history engine. Baseline rates of major GI bleeding in the nonelevated risk population (that is, among persons for whom aspirin use is not contraindicated) were estimated by using data from a large Italian population-based cohort study (47), with adjustments for the U.S. age and sex distribution (Appendix Table 3). Case-fatality rates for GI bleeding, based on patients without complicating comorbidities, were derived from a prospective study conducted in the United Kingdom (48). Baseline CRC incidence rates used in the model are derived from U.S. data (49, 50) and reflect contemporary use of screening technologies, such as colonoscopy, which can prevent CRC by the identification and removal of cancer precursors.

Appendix Table 2. Comparison of Baseline Modeled CVD Event Rates With National Prevalence Estimates*

Image: M152129tt5_Appendix_Table_2_Comparison_of_Baseline_Modeled_CVD_Event_Rates_With_National_Prevale

Appendix Table 2. Comparison of Baseline Modeled CVD Event Rates With National Prevalence Estimates*

Appendix Table 3. Baseline GI Bleeding and Case-Fatality Rate Parameter Values*

Uncertainty and Sensitivity Analysis

Two sources of uncertainty were considered in this study: stochastic heterogeneity from the variability in outcomes experienced by a randomly selected sample population and parameter uncertainty from the imprecision of model parameter estimates (51). Confidence intervals reflecting stochastic heterogeneity were estimated by bootstrap resampling the simulated population for each stratified outcome 100 000 times with replacement. Deterministic (1-way) sensitivity analyses of key parameters were conducted with all other parameters, probabilities, and population characteristics held equal. Table 1 and Appendix Tables 1 and 3 present the alternative parameter values used in the deterministic analyses. Probabilistic sensitivity analyses can be found in our prior work (52).

Role of Funding Source

The Agency for Healthcare Research and Quality provided funding, project oversight, and review for this study. Four USPSTF members helped to resolve scope and methodological issues, and 4 peer reviewers provided feedback on draft findings. Final model results are the sole responsibility of the authors.

Results

Lifetime Net Benefit

The estimated lifetime net difference in QALYs from using aspirin for primary prevention is positive for all sex and baseline CVD risk groups aged 40 to 69 years (range, 7.4 to 107.9 QALYs per 1000 persons) that we considered (Table 2). In our results, net life-years are positive for nearly all groups aged 40 to 59 years (range, 3.2 to 82.8 life-years per 1000 persons). For women aged 50 to 59 years with a 10-year CVD risk of 1% and both sexes aged 60 to 69 years with a 10-year CVD risk of 10% or less, net life-years are negative. Both net QALYs and life-years are negative for men and women of all considered risk levels aged 70 to 79 years. The magnitude of lifetime net life-years and QALYs is often similar for men and women and is generally greater the lower the age or the greater the 10-year CVD risk at initiation.

Table 2. Net Life-Years and QALYs of Lifetime, 20-y, and 10-y Aspirin Use*

Detailed benefit and harm outcomes are presented in Table 3 and Appendix Tables 4 and 5. Differences in lifetime net outcomes between men and women are explained by the differences in baseline incidence for MI (higher for men), ischemic stroke (higher for women), and GI bleeding (higher for men). Women also have a longer life expectancy, which corresponds to a longer average risk exposure during which aspirin can intervene. When comparing by age groups, we found that lifetime net CVD events and prevented CRC cases are at their lowest when aspirin is initiated at older ages. This corresponds to the decrease in person-years of risk exposure. In contrast, lifetime net harms are similar or greater among older age groups because of increases in baseline GI bleeding and hemorrhagic stroke risk with age. Persons with lower CVD risk often have greater expected reductions in CRC incidence and mortality because of longer life expectancy. Because of the complex interplay between benefits and harms of aspirin on the length or quality of life, the sign of net events does not always correspond with net life-years or net QALYs.

Table 3. Detailed Benefit and Harm Tradeoffs of Aspirin Use With a CVD Risk of 10%*

Appendix Table 4. Expanded Lifetime Benefit and Harm Tradeoffs of Aspirin Use for Men Aged 40–79 y*

Image: M152129tt7_Appendix_Table_4_Expanded_Lifetime_Benefit_and_Harm_Tradeoffs_of_Aspirin_Use_for_Men

Appendix Table 4. Expanded Lifetime Benefit and Harm Tradeoffs of Aspirin Use for Men Aged 40–79 y*

Appendix Table 5. Expanded Lifetime Benefit and Harm Tradeoffs of Aspirin Use for Women Aged 40–79 y*

Image: M152129tt8_Appendix_Table_5_Expanded_Lifetime_Benefit_and_Harm_Tradeoffs_of_Aspirin_Use_for_Wome

Appendix Table 5. Expanded Lifetime Benefit and Harm Tradeoffs of Aspirin Use for Women Aged 40–79 y*

Net Benefit Over 10 and 20 Years

Over 20 years, the predicted net QALYs from aspirin remain positive for most CVD risk groups (men and women) aged 40 to 69 years (Table 2). However, the magnitude of net QALYs that is positive over 20 years is generally a small fraction of the lifetime net benefit (range, 0.1 to 23.6 QALYs per 1000 persons). In addition, net life-years are negative for nearly all groups over this time frame. Over 10 years, net life-years and QALYs are also negative, or are only marginally positive, for all groups. No CRC benefit is reflected in the 10-year results because of the delayed effect found in the systematic evidence review.

Sensitivity Analyses

One-way parameter sensitivity for men and women with baseline CVD risk of 10% over lifetime, 20-year, and 10-year horizons are compared in Appendix Tables 6, 7 and 8, respectively. These tables show that the possibility for a direct reduction in the RR for CVD-related death from aspirin (cases 6 and 7) has the most potential to sway results; net life-years and QALYs would be positive for nearly all groups over all time horizons with only a 3% reduction in CVD mortality risk (case 7). The next most sensitive parameter to both measures is the RR for hemorrhagic stroke (cases 13 and 14). Even a small disutility associated with taking aspirin routinely (cases 1 and 2) can dramatically decrease net QALYs. Aspirin's effect on reducing CRC incidence also has a considerable effect; not accounting for this effect reduces lifetime net QALYs by about 50% and lifetime net life-years by even more. Variation in GI bleeding incidence and case-fatality rates (cases 10 to 12) has a greater relative effect for persons initiating aspirin in their 60s and 70s.

Appendix Table 6. Comparisons in Lifetime Net Benefit of Taking Aspirin for Men and Women With a CVD Risk of 10%*

Image: M152129tt9_Appendix_Table_6_Comparisons_in_Lifetime_Net_Benefit_of_Taking_Aspirin_for_Men_and_Wo

Appendix Table 6. Comparisons in Lifetime Net Benefit of Taking Aspirin for Men and Women With a CVD Risk of 10%*

Appendix Table 7. Comparisons in Net Benefit of Taking Aspirin Over 20 y for Men and Women With a CVD Risk of 10%*

Image: M152129tt10_Appendix_Table_7_Comparisons_in_Net_Benefit_of_Taking_Aspirin_Over_20_y_for_Men_and

Appendix Table 7. Comparisons in Net Benefit of Taking Aspirin Over 20 y for Men and Women With a CVD Risk of 10%*

Appendix Table 8. Comparisons in Net Benefit of Taking Aspirin Over 10 y for Men and Women With a CVD Risk 10%*

Image: M152129tt11_Appendix_Table_8_Comparisons_in_Net_Benefit_of_Taking_Aspirin_Over_10_y_for_Men_and

Appendix Table 8. Comparisons in Net Benefit of Taking Aspirin Over 10 y for Men and Women With a CVD Risk 10%*

Discussion

These estimates quantify the expected difference in benefits and harms from taking low-dose aspirin for the primary prevention of CVD and CRC by age, sex, and baseline 10-year CVD risk group as derived from a detailed microsimulation model. Overall, we find that aspirin is expected to improve overall quality of life (that is, reduce illness) for most men and women without elevated bleeding risk when aspirin is initiated at ages 40 to 69 years for lifetime use, unless otherwise contraindicated. Such use is also expected to improve life expectancy for most men and women who start aspirin at ages 40 to 59 years and for those at higher CVD risk who start aspirin at ages 60 to 69 years. Our primary results do not find overall benefits to outweigh harms for persons in their 70s with a 10-year CVD risk of 20% or less. The balance of benefits and harms from using aspirin over 10 and 20 years is far more tenuous for most population groups, and several limitations and considerations should be considered before translating any of these findings to practice.

This study incorporates important new evidence that has been published since the last USPSTF reviews (3, 4). One major difference is in our findings by sex. Aspirin was previously found to reduce the RR for MI in men by 32% and the RR for stroke in women by 17%; the current review finds that aspirin reduces the RR for MI by 17% and the RR for stroke by 14% (12) in both men and women. Another major difference is the new finding of lower risk for CRC after 10 years of aspirin use. This added benefit can account for more than half of the lifetime net benefit, in terms of life-years and QALYs, from routine aspirin use (case 3 in Appendix Table 6). Of note, the RR for GI bleeding with aspirin was previously 2.00 compared with 1.58 found in the updated review. The RR for hemorrhagic stroke was previously 1.69 compared with the substantially lower RR of 1.27 used in this study.

We also had many methodological differences (detailed in section 5.1 of the Supplement). The prior net benefit calculations were restricted to the first nonfatal events over 10 years (Table 24 of the Supplement). In this analysis, we account for fatal and nonfatal events over a lifetime and provide life-years and QALYs as outcome measures. Our results reveal that the lifetime horizon is needed to ensure all important benefits and harms are captured and that the largest average net balance of benefits is realized with long-term aspirin use. Life-years are an important measure because they incorporate differences in the expected length of life that may come from increased prevalence of fatal bleeding episodes, which are balanced against indirect reductions in CVD or CRC mortality that arise from the prevention of nonfatal CVD and CRC incidence. Quality-adjusted life-years are an important measure because they incorporate both expected length- and quality-of-life effects, which balance all fatal and nonfatal benefits harms. In addition, the ratio of nonfatal to fatal events generally decreases with age; therefore, we find fewer preventable nonfatal MIs and ischemic strokes in older age groups in our competing risk framework. Calculations of harms also differ. Our analysis incorporates estimates of age-adjusted case-fatality associated with GI bleeding events. This can have a meaningful effect on net benefit calculations, particularly for men and women initiating aspirin in their 60s and 70s (case 10 in Appendix Table 6). In addition, hemorrhagic stroke rates vary by age and CVD risk groups, which means that both benefits and harms scale with baseline CVD risk in our analysis. The baseline hemorrhagic stroke rates generated by our model compare well with large U.S.-based cohort studies and are generally much higher than assumed by the 2009 recommendation (Table 25 of the Supplement).

A recent study (53) used long-term follow-up results from the Women's Health Study to develop competing risk prediction models to estimate absolute risk reduction among CVD, cancer, and GI bleeding. Despite differences in underlying evidence and methods, our results over 10 and approximately 15 years are generally similar (Table 27 of the Supplement). Another recent study (54) used a population-based incidence model to estimate the net difference in event rates over 15 years with prophylactic aspirin use in the general U.K. population. For that which can be compared, findings are generally consistent between studies, with differences in net events attributable to differential baseline event rates between the U.S. and U.K. populations and by the combined-versus-separated approach to aspirin's effect on stroke type (Table 28 of the Supplement). In addition, a cohort modeling study (55) examined the cost-effectiveness of aspirin for primary prevention of CVD and CRC in men. Findings are again generally similar, with differences in net QALYs largely explained by the inclusion of a disutility to taking aspirin in the study's base-case analysis.

The average effectiveness of aspirin is determined by randomized trials and may reflect cross-contamination if participants assigned to the control group chose to use aspirin or those assigned to the aspirin group chose not to use aspirin (nonadherence). It is not known what would be observed with typical adherence levels and a pure control group. We expect, however, that the effectiveness of aspirin reflected in our analysis should correspond with good adherence because the modeled population mirrors persons willing to participate in a clinical trial. Our findings may not be relevant to those with lower expected adherence patterns, especially if they are associated with disutility for taking aspirin.

The systematic reviews did not find compelling evidence of differential effects by age. It is not clear how robust the homogenous RR effects are for all population groups, particularly persons with low event rates in the trial populations (such as those in their early 40s). We extended aspirin effects for persons older than 80 years; however, we did not evaluate aspirin initiation for these persons (nor for those <40 years) because of limited representation at enrollment in aspirin trials. Results from the ongoing ASPREE (Aspirin in Reducing Events in the Elderly) trial (56) may help to fill data gaps among older populations.

It is widely believed that aspirin reduces the risk for ischemic stroke but increases the risk for hemorrhagic stroke. The latter was not found to be statistically significant in the systematic reviews, but we included this harm in our decision analysis because of its biological plausibility and the limited power to detect differences in this relatively rare event in study populations. In addition, our RR estimate for ischemic stroke underestimates benefits because it is derived from combined stroke data. This conservative approach may be appropriate given the imprecision in measuring the increased risk for hemorrhagic stroke.

By design, both CVD and CRC mortality risk may be affected indirectly by aspirin use in our model as a downstream effect of preventing nonfatal CVD events or CRC incidence, respectively. Low-dose aspirin trials indicate that there may be a small reduction in CVD mortality risk, but this finding is not statistically significant (12) and we did not include it in our base-case analysis. In contrast, evidence indicates that the RRs for CRC incidence and mortality are both reduced with aspirin use (13). To avoid double-counting, we modeled aspirin's effect on CRC incidence only and allowed the model's natural history of cancer to determine CRC deaths prevented. Nevertheless, we found that CVD mortality at 10 years and CRC mortality at 20 years in our model results align with rates observed in trials (Table 29 of the Supplement).

The model accounts for a correlation between risk for CVD and CRC because of tobacco use. Hemorrhagic stroke risk also correlates with overall CVD risk. We did not, however, establish and incorporate GI bleeding risk equations that would account for a correlation between GI bleeding and CVD risk factors, such as tobacco use and diabetes.

These results naturally raise questions about whether there is an optimal age to stop aspirin use; however, evidence is lacking on the implications of discontinuing aspirin after long-term use. It could be misleading to use a model to inform discontinuation decisions without better data to support such analyses.

This analysis approached the decision to use aspirin from the perspective of a person's age, sex, and 10-year risk for CVD. Given the systematic evidence review findings of substantial benefit from aspirin in the prevention of CRC, persons with an elevated risk for CRC may consider using aspirin for this benefit alone. Stratifying net benefits by CRC risk was outside the scope of this analysis, but the detailed outcomes presented in Table 3 and Appendix Tables 4 and 5 may be helpful for those considering aspirin use for that reason.

These results apply to persons whose aspirin use is not absolutely contraindicated by a medical provider for such reasons as a history of GI or intracranial bleeding or concurrent use of medications that increase bleeding risk, which corresponds with the data we used to inform community-based bleeding risks (47). We did not account for heterogeneity in GI bleeding risks beyond age and sex. Case-fatality rates for GI bleeding events are not well-established in the literature. Aspirin primary prevention trials do not show a difference in GI bleeding mortality, but this may be due to the rarity of these events in highly selected trial populations. Our analysis uses observational GI bleeding mortality data (48), which indicate a large increase in case-fatality rates at older ages. Better estimates of how age, sex, aspirin, and other possible risk factors interact to affect GI bleeding and case-fatality rates may modify the net benefit findings, especially among older age groups (cases 10 to 12 in Appendix Tables 6, 7 and 8).

These results indicate that several population groups may benefit from aspirin for the primary prevention of CVD and CRC. Nevertheless, discretion should be used when interpreting these results because sensitivity analyses reveal meaningful uncertainty about the magnitude of net benefit. Benefit and harm calculations are most sensitive to uncertainty about the effect of low-dose aspirin on the increased risk for hemorrhagic stroke and in the primary prevention of CVD mortality. Moreover, parameter estimates used in this study may not be reliable for populations underrepresented in the aspirin primary prevention trials. A better understanding of the effects of aspirin by age group and after discontinuation, additional studies that report aspirin's effect on ischemic stroke separately from hemorrhagic stroke, and the development of comprehensive risk equations for GI bleeding would increase confidence in and precision of the simulation results. Quality-of-life benefits from using aspirin may be considerably diminished among persons who dislike taking routine medications. Future research may identify additional benefits (such as protective effects against other types of cancer) or harms that may substantially alter these findings. These sources of uncertainty and patient preferences should be carefully considered in the shared decision-making process about the routine use of aspirin for primary prevention.

References

Antithrombotic Trialists' Collaboration
Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients.
BMJ
2002
324
71
86
PubMed
CrossRef

PubMed
Baigent
C
Blackwell
L
Collins
R
Emberson
J
Godwin
J
Peto
R
et al
Antithrombotic Trialists' (ATT) Collaboration.
Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials.
Lancet
2009
373
1849
60
PubMed
CrossRef

PubMed
U.S. Preventive Services Task Force
Routine aspirin or nonsteroidal anti-inflammatory drugs for the primary prevention of colorectal cancer: U.S. Preventive Services Task Force recommendation statement.
Ann Intern Med
2007
146
361
4
CrossRef

PubMed
U.S. Preventive Services Task Force
Aspirin for the prevention of cardiovascular disease: U.S. Preventive Services Task Force recommendation statement.
Ann Intern Med
2009
150
396
404
CrossRef

PubMed
Berger
JS
Roncaglioni
MC
Avanzini
F
Pangrazzi
I
Tognoni
G
Brown
DL
Aspirin for the primary prevention of cardiovascular events in women and men: a sex-specific meta-analysis of randomized controlled trials.
JAMA
2006
295
306
13
PubMed
CrossRef

PubMed
Krantz
MJ
Berger
JS
Hiatt
WR
An aspirin a day: are we barking up the wrong willow tree?[Editorial].
Pharmacotherapy
2010
30
115
8
PubMed
CrossRef

PubMed
Gaziano
JM
Greenland
P
When should aspirin be used for prevention of cardiovascular events? [Editorial].
JAMA
2014
312
2503
4
PubMed
CrossRef

PubMed
Pignone
M
Aspirin for primary prevention: a challenging decision [Editorial].
J Am Heart Assoc
2014
3
PubMed
Guirguis-Blake
JM
Evans
CV
Senger
CA
Rowland
MG
O'Connor
EA
Whitlock
EP
Aspirin for the Primary Prevention of Cardiovascular Events: A Systematic Evidence Review for the U.S. Preventive Services Task Force. Evidence Synthesis No. 131. AHRQ Publication No. 13-05195-EF-1.
Rockville, MD
Agency for Healthcare Research and Quality
2015
Chubak
J
Kamineni
A
Buist
DSM
Anderson
ML
Whitlock
EP
Aspirin Use for the Prevention of Colorectal Cancer: An Updated Systematic Evidence Review for the U.S. Preventive Services Task Force. Evidence Synthesis No. 133. AHRQ Publication No. 15-05228-EF-1.
Rockville, MD
Agency for Healthcare Research and Quality
2015
Whitlock
EP
Williams
SB
Burda
BU
Feightner
A
Beil
T
Aspirin Use in Adults: Cancer, All-Cause Mortality and Harms. A Systematic Evidence Review for the U.S. Preventive Services Task Force. Evidence Synthesis No. 132. AHRQ Publication No. 13-05193-EF-1.
Rockville, MD
Agency for Healthcare Research and Quality
2015
Guirguis-Blake
JM
Evans
CV
Senger
CA
O'Connor
EA
Whitlock
EP
Aspirin for the primary prevention of cardiovascular events: a systematic evidence review for the U.S. Preventive Services Task Force.
Ann Intern Med
2016
164
804
13
CrossRef

PubMed
Chubak
J
Whitlock
EP
Williams
SB
Kamineni
A
Burda
BU
Buist
DSM
et al
Aspirin for the prevention of cancer incidence and mortality: systematic evidence reviews for the U.S. Preventive Services Task Force.
Ann Intern Med
2016
164
814
25
CrossRef

PubMed
Whitlock
EP
Burda
BU
Williams
SB
Guirguis-Blake
JM
Evans
CV
Bleeding risks with aspirin use for primary prevention in adults: a systematic review for the U.S. Preventive Services Task Force.
Ann Intern Med
2016
164
826
35
CrossRef

PubMed
Bibbins-Domingo
K
U.S. Preventive Services Task Force
Aspirin use for the primary prevention of cardiovascular disease and colorectal cancer: U.S. Preventive Services Task Force recommendation statement.
Ann Intern Med
2016
164
836
45
CrossRef

PubMed
Goff
DC
Jr
Lloyd-Jones
DM
Bennett
G
Coady
S
D'Agostino
RB
Sr
Gibbons
R
et al
American College of Cardiology/American Heart Association Task Force on Practice Guidelines
2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines.
J Am Coll Cardiol
2014
63
2935
59
PubMed
CrossRef

PubMed
Annual estimates of the resident population by sex, race alone, and Hispanic origin for counties: April 1, 2010 to July 1, 2011.
Washingon, DC
U.S. Census Bureau
2011
Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey: NHANES 2001-2002. Hyattsville, MD: National Center for Health Statistics; 2004. Accessed at wwwn.cdc.gov/nchs/nhanes/search/nhanes01_02.aspx on 14 March 2016.
Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey: NHANES 2003-2004. Hyattsville, MD: National Center for Health Statistics; 2005. Accessed at wwwn.cdc.gov/Nchs/Nhanes/Search/nhanes03_04.aspx on 14 March 2016.
Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey: NHANES 2005-2006. Hyattsville, MD: National Center for Health Statistics; 2007. Accessed at wwwn.cdc.gov/Nchs/Nhanes/Search/nhanes05_06.aspx on 14 March 2016.
Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey: NHANES 2007-2008. Hyattsville, MD: National Center for Health Statistics; 2009. Accessed at wwwn.cdc.gov/Nchs/Nhanes/Search/nhanes07_08.aspx on 14 March 2016.
Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey: NHANES 2009-2010. Hyattsville, MD: National Center for Health Statistics; 2011. Accessed at wwwn.cdc.gov/Nchs/Nhanes/Search/nhanes09_10.aspx on 14 March 2016.
National Center for Health Statistics. National Health Interview Survey: 2007 data release. Hyattsville, MD: National Center for Health Statistics; 2008. Accessed at www.cdc.gov/nchs/nhis/nhis_2007_data_release.htm on 14 March 2016.
Congressional Budget Office. Raising the excise tax on cigarettes: effects on health and the federal budget. Washington, DC: Congressional Budget Office; 2012. Accessed at www.cbo.gov/publication/43319 on 14 March 2016.
Cost-effectiveness in Health and Medicine.
New York
Oxford Univ Pr
1996
Wolff
T
Miller
T
Ko
S
Aspirin for the primary prevention of cardiovascular events: an update of the evidence for the U.S. Preventive Services Task Force.
Ann Intern Med
2009
150
405
10
CrossRef

PubMed
Fowkes
FG
Price
JF
Stewart
MC
Butcher
I
Leng
GC
Pell
AC
et al
Aspirin for Asymptomatic Atherosclerosis Trialists
Aspirin for prevention of cardiovascular events in a general population screened for a low ankle brachial index: a randomized controlled trial.
JAMA
2010
303
841
8
PubMed
CrossRef

PubMed
Hansson
L
Zanchetti
A
Carruthers
SG
Dahlöf
B
Elmfeldt
D
Julius
S
et al
Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial.
HOT Study Group. Lancet
1998
351
1755
62
PubMed
Ogawa
H
Nakayama
M
Morimoto
T
Uemura
S
Kanauchi
M
Doi
N
et al
Japanese Primary Prevention of Atherosclerosis With Aspirin for Diabetes (JPAD) Trial Investigators
Low-dose aspirin for primary prevention of atherosclerotic events in patients with type 2 diabetes: a randomized controlled trial.
JAMA
2008
300
2134
41
PubMed
CrossRef

PubMed
Belch
J
MacCuish
A
Campbell
I
Cobbe
S
Taylor
R
Prescott
R
et al
Prevention of Progression of Arterial Disease and Diabetes Study Group
The Prevention Of Progression of Arterial Disease And Diabetes (POPADAD) trial: factorial randomised placebo controlled trial of aspirin and antioxidants in patients with diabetes and asymptomatic peripheral arterial disease.
BMJ
2008
337
a1840
.
PubMed
CrossRef

PubMed
de Gaetano
G
Collaborative Group of the Primary Prevention Project
Low-dose aspirin and vitamin E in people at cardiovascular risk: a randomised trial in general practice. Collaborative Group of the Primary Prevention Project.
Lancet
2001
357
89
95
PubMed
CrossRef

PubMed
Thrombosis prevention trial: randomised trial of low-intensity oral anticoagulation with warfarin and low-dose aspirin in the primary prevention of ischaemic heart disease in men at increased risk. The Medical Research Council's General Practice Research Framework.
Lancet
1998
351
233
41
PubMed
CrossRef

PubMed
Ridker
PM
Cook
NR
Lee
IM
Gordon
D
Gaziano
JM
Manson
JE
et al
A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women.
N Engl J Med
2005
352
1293
304
PubMed
CrossRef

PubMed
Ikeda
Y
Shimada
K
Teramoto
T
Uchiyama
S
Yamazaki
T
Oikawa
S
et al
Low-dose aspirin for primary prevention of cardiovascular events in Japanese patients 60 years or older with atherosclerotic risk factors: a randomized clinical trial.
JAMA
2014
312
2510
20
PubMed
CrossRef

PubMed
Cook
NR
Lee
IM
Zhang
SM
Moorthy
MV
Buring
JE
Alternate-day, low-dose aspirin and cancer risk: long-term observational follow-up of a randomized trial.
Ann Intern Med
2013
159
77
85
CrossRef

PubMed
Flossmann
E
Rothwell
PM
British Doctors Aspirin Trial and the UK-TIA Aspirin Trial
Effect of aspirin on long-term risk of colorectal cancer: consistent evidence from randomised and observational studies.
Lancet
2007
369
1603
13
PubMed
CrossRef

PubMed
Salomon
JA
Vos
T
Hogan
DR
Gagnon
M
Naghavi
M
Mokdad
A
et al
Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010.
Lancet
2012
380
2129
43
PubMed
CrossRef

PubMed
Nyman
JA
Barleen
NA
Dowd
BE
Russell
DW
Coons
SJ
Sullivan
PW
Quality-of-life weights for the US population: self-reported health status and priority health conditions, by demographic characteristics.
Med Care
2007
45
618
28
PubMed
CrossRef

PubMed
Sullivan
PW
Lawrence
WF
Ghushchyan
V
A national catalog of preference-based scores for chronic conditions in the United States.
Med Care
2005
43
736
49
PubMed
CrossRef

PubMed
Mittmann
N
Trakas
K
Risebrough
N
Liu
BA
Utility scores for chronic conditions in a community-dwelling population.
Pharmacoeconomics
1999
15
369
76
PubMed
CrossRef

PubMed
Fryback
DG
Dasbach
EJ
Klein
R
Klein
BE
Dorn
N
Peterson
K
et al
The Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors.
Med Decis Making
1993
13
89
102
PubMed
CrossRef

PubMed
Gold MR, Siegel JE, Russell LB, Weinstein MC, eds. Cost-effectiveness in Health and Medicine. New York: Oxford Univ Pr 1996.
Sullivan
PW
Ghushchyan
V
Preference-Based EQ-5D index scores for chronic conditions in the United States.
Med Decis Making
2006
26
410
20
PubMed
CrossRef

PubMed
Ko
CY
Maggard
M
Livingston
EH
Evaluating health utility in patients with melanoma, breast cancer, colon cancer, and lung cancer: a nationwide, population-based assessment.
J Surg Res
2003
114
1
5
PubMed
CrossRef

PubMed
Djalalov
S
Rabeneck
L
Tomlinson
G
Bremner
KE
Hilsden
R
Hoch
JS
A review and meta-analysis of colorectal cancer utilities.
Med Decis Making
2014
34
809
818
PubMed
CrossRef

PubMed
Lee
HY
Hwang
JS
Jeng
JS
Wang
JD
Quality-adjusted life expectancy (QALE) and loss of QALE for patients with ischemic stroke and intracerebral hemorrhage: a 13-year follow-up.
Stroke
2010
41
739
44
PubMed
CrossRef

PubMed
De Berardis
G
Lucisano
G
D'Ettorre
A
Pellegrini
F
Lepore
V
Tognoni
G
et al
Association of aspirin use with major bleeding in patients with and without diabetes.
JAMA
2012
307
2286
94
PubMed
CrossRef

PubMed
Rockall
TA
Logan
RF
Devlin
HB
Northfield
TC
Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom.
BMJ
1995
311
222
6
PubMed
CrossRef

PubMed
National Cancer Institute; Surveillance Research Program
SEER*Stat, version 8.1.5.
Bethesda, MD
National Cancer Institute
2014
Centers for Disease Control and Prevention
Smoking-Attributable Mortality, Morbidity, and Economic Costs (SAMMEC): Adult SAMMEC and Maternal and Child Health (MCH) SAMMEC software.
Atlanta, GA
Centers for Disease Control and Prevention
2007
Briggs
AH
Weinstein
MC
Fenwick
EA
Karnon
J
Sculpher
MJ
Paltiel
AD
ISPOR-SMDM Modeling Good Research Practices Task Force
Model parameter estimation and uncertainty analysis: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force Working Group-6.
Med Decis Making
2012
32
722
32
PubMed
CrossRef

PubMed
Dehmer
SP
Maciosek
MV
Flottemesch
TJ
Aspirin Use to Prevent Cardiovascular Disease and Colorectal Cancer: A Decision Analysis: Evidence Synthesis No 131. AHRQ Publication No. 15-05229-EF-1.
Rockville, MD
Agency for Healthcare Research and Quality
2015
van Kruijsdijk
RC
Visseren
FL
Ridker
PM
Dorresteijn
JA
Buring
JE
van der Graaf
Y
et al
Individualised prediction of alternate-day aspirin treatment effects on the combined risk of cancer, cardiovascular disease and gastrointestinal bleeding in healthy women.
Heart
2015
101
369
76
PubMed
CrossRef

PubMed
Cuzick
J
Thorat
MA
Bosetti
C
Brown
PH
Burn
J
Cook
NR
et al
Estimates of benefits and harms of prophylactic use of aspirin in the general population.
Ann Oncol
2015
26
47
57
PubMed
CrossRef

PubMed
Pignone
M
Earnshaw
S
McDade
C
Pletcher
MJ
Effect of including cancer mortality on the cost-effectiveness of aspirin for primary prevention in men.
J Gen Intern Med
2013
28
1483
91
PubMed
CrossRef

PubMed
Nelson
M
Reid
C
Beilin
L
Donnan
G
Johnston
C
Krum
H
et al
Aspirin in Reducing Events in the Elderly (ASPREE) Study Group
Rationale for a trial of low-dose aspirin for the primary prevention of major adverse cardiovascular events and vascular dementia in the elderly: Aspirin in Reducing Events in the Elderly (ASPREE).
Drugs Aging
2003
20
897
903
PubMed
CrossRef

PubMed