An accurate assessment of drug safety or effectiveness in pharmaco-epidemiology requires defining an etiologically correct time-varying exposure model, which specifies how previous drug use affects the hazard of the event of interest. An additional challenge is to account for the multitude of mutually exclusive events that may be associated with the use of a given drug. To simultaneously address both challenges, we develop, and validate in simulations, a new approach that combines flexible modeling of the cumulative effects of time-varying exposures with competing risks methodology to separate the effects of the same drug exposure on different outcomes. To account for the dosage, duration and timing of past exposures, we rely on a spline-based weighted cumulative exposure modeling. We also propose likelihood ratio tests to test if the cumulative effects of past exposure on the hazards of the competing events are the same or different. Simulation results indicate that the estimated event-specific weight functions are reasonably accurate, and that the proposed tests have acceptable type I error rate and power. In real-life application, the proposed method indicated that recent use of antihypertensive drugs may reduce the risk of stroke but has no effect on the hazard of coronary heart disease events.
AvornJ. In defense of pharmacoepidemiology – embracing the yin and yang of drug research. N Engl J Med2007; 357: 2219–2221.
2.
AbrahamowiczMTamblynRDrug utilization patterns. In: ArmitagePColtonT (eds). Encyclopedia of biostatistics2005; Vol. 4, 2nd ed. Chichester: John Wiley & Sons, Ltd, pp. 1533–1553.
3.
PatornoEGarryEMPatrickARet al.Addressing limitations in observational studies of the association between glucose-lowering medications and all-cause mortality: a review. Drug Saf2015; 38: 295–310.
4.
AllignolABeyersmannJSchmoorC. Statistical issues in the analysis of adverse events in time-to-event data. Pharm Stat2016; 15: 297–305.
5.
ProctorTSchumacherM. Analysing adverse events by time-to-event models: the CLEOPATRA study. Pharm Stat2016; 15: 306–314.
6.
ColletJPSharpeCBelzileEet al.Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: effects of dosage and timing. Br J Cancer1999; 81: 62–68.
7.
GambleJMSimpsonSHEurichDTet al.Insulin use and increased risk of mortality in type 2 diabetes: a cohort study. Diabetes Obes Metab2010; 12: 47–53.
8.
LacailleDGuhDPAbrahamowiczMet al.Use of nonbiologic disease-modifying antirheumatic drugs and risk of infection in patients with rheumatoid arthritis. Arthritis Care Res2008; 59: 1074–1081.
9.
SmittenALHochbergMCSuissaSet al.The risk of hospitalized infection in patients with rheumatoid arthritis. J Rheumatol2008; 35: 387–393.
10.
DixonWGAbrahamowiczMBeauchampMEet al.Immediate and delayed impact of oral glucocorticoid therapy on risk of serious infection in patients with rheumatoid arthritis: a nested case-control analysis. Ann Rheum Dis2012; 71: 1128–1133.
11.
RadawskiCMorratoEHornbuckleKet al.on behalf of the BRACE Special Interest Group. Benefit–Risk Assessment, Communication, and Evaluation (BRACE) throughout the life cycle of therapeutic products: overall perspective and role of the pharmacoepidemiologist. Pharmacoepidemiol Drug Saf2015; 24: 1233–1240.
12.
AbrahamowiczMBeauchampMESylvestreMP. Comparison of alternative models for linking drug exposure with adverse effects. Stat Med2012; 31: 1014–1030.
13.
HauptmannMWellmannJLubinJHet al.Analysis of exposure-time-response relationships using a spline weight function. Biometrics2000; 56: 1105–1108.
14.
VacekPM. Assessing the effect of intensity when exposure varies over time. Stat Med1997; 16: 505–513.
15.
JickHGarcia RodriguezLAPerez-GutthannS. Principles of epidemiological research on adverse and beneficial drug effects. Lancet1998; 352: 1767–1770.
16.
SkeggDCG. Evaluating the safety of medicines, with particular reference to contraception. Stat Med2001; 20: 3557–3569.
17.
XiaoYAbrahamowiczMMoodieEEMet al.Flexible marginal structural models for estimating the cumulative effect of a time-dependent treatment on the hazard: reassessing the cardiovascular risks of didanosine treatment in the Swiss HIV cohort study. J Am Stat Assoc2014; 109: 455–464.
18.
van GaalenRDAbrahamowiczMBuckeridgeDL. The impact of exposure model misspecification on signal detection in prospective pharmacovigilance. Pharmacoepidemiol Drug Saf2015; 24: 456–467.
19.
StrangesSBonnerMRFucciFet al.Lifetime cumulative exposure to secondhand smoke and risk of myocardial infarction in never smokers: results from the Western New York health study, 1995–2001. Arch Intern Med2006; 166: 1961–1967.
20.
van LeeuwenRVingerlingJRHofmanAet al.Cholesterol lowering drugs and risk of age related maculopathy: prospective cohort study with cumulative exposure measurement. Br Med J2003; 26: 255–256.
21.
BreslowNELubinJHMarekPet al.Multiplicative models and cohort analysis. J Am Stat Soc1983; 78: 1–12.
22.
ThomasDC. Models for exposure-time-response relationships with applications to cancer epidemiology. Ann Rev Public Health1988; 9: 451–482.
23.
EdwardsRAronsonJ. Adverse drug reactions: definitions, diagnosis, and management. Lancet2000; 356: 1255–1259.
24.
GrimJChládekJMartínkováJ. Pharmacokinetics and pharmacodynamics of methotrexate in non-neoplastic diseases. Clin Pharmacokinet2003; 42: 139–151.
25.
AbrahamowiczMBartlettGTamblynRet al.Modeling cumulative dose and exposure duration provided insights regarding the associations between benzodiazepines and injuries. J Clin Epidemiol2006; 59: 393–403.
26.
SylvestreMPAbrahamowiczM. Flexible modeling of the cumulative effects of time-dependent exposures on the hazard. Stat Med2009; 28: 3437–3453.
27.
SylvestreMPAbrahamowiczMCapekRet al.Assessing the cumulative effects of exposure to selected benzodiazepines on the risk of fall-related injuries in the elderly. Int Psychogeriatr2012; 24: 577–586.
28.
YoungJXiaoYMoodieEEMet al.and the Swiss HIV Cohort Study. Effect of cumulating exposure to abacavir on the risk of cardiovascular disease events in patients from the Swiss HIV cohort study. J Acquir Immune Defic Syndr2015; 69: 413–421.
29.
MovahediMBeauchampMEAbrahamowiczMet al.Risk of incident diabetes mellitus associated with the dosage and duration of oral glucocorticoid therapy in patients with rheumatoid arthritis. Arthritis Rheumatol2016; 68: 1089–1098.
30.
YuDPeatGBedsonJet al.Weighted cumulative exposure models helped identify an association between early knee-pain consultations and future knee OA diagnosis. J Clin Epidemiol2016; 76: 218–228.
31.
International drug monitoring: the role of national centres. Report of WHO meeting. World Health Organization. 1972.
32.
NissenSEWolskiKTopolEJ. Effect of muraglitazar on death and major adverse cardiovascular events in patients with type 2 diabetes mellitus. J Am Med Assoc2005; 294: 2581–2586.
33.
AldermanMHDavisBRPillerLBet al.ALLHAT Collaborative Research Group. Should antihypertensive treatment recommendations differ in patients with and without coronary heart disease? (from the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial [ALLHAT]). Am J Cardiol2016; 117: 105–115.
34.
LimEBrownAHelmyAet al.Composite outcomes in cardiovascular research: a survey of randomized trials. Ann Intern Med2008; 149: 612–617.
35.
Ferreira-GonzálezIBusseJWHeels-AnsdellDet al.Problems with use of composite end points in cardiovascular trials: systematic review of randomized controlled trials. BMJ2007; 334: 786–786.
36.
HoltJD. Competing risks analyses with special reference to matched pair experiments. Biometrika1978; 65: 159–165.
37.
PutterHFioccoMGeskusRB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med2007; 26: 2389–2430.
38.
AndersenPKBorganO. Counting process models for life history data. A review. Scand J Stat1985; 12: 97–158.
39.
PrenticeRLKalbfleischJDPetersonAV.Jret al.The analysis of failure times in the presence of competing risks. Biometrics1978; 34: 541–554.
40.
PutterHFioccoMGeskusRB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med2007; 26: 2389–2430.
41.
LunnMMcNeilD. Applying Cox regression to competing risks. Biometrics1995; 51(2): 524–532.
42.
BelotAAbrahamowiczMRemontetLet al.Flexible modeling of competing risks in survival analysis. Stat Med2010; 29: 2453–2468.
43.
AbrahamowiczMMacKenzieTEsdaileJM. Time-dependent hazard ratio: modeling and hypothesis testing with application in lupus nephritis. J Am Stat Assoc1996; 91: 1432–1439.
44.
AkaikeH. A new look at statistical model identification. IEEE Trans Automatic Control1974; 19: 716–723.
45.
RachetBAbrahamowiczMSascoAJet al.Estimating the distribution of lag in the effect of short-term exposures and interventions: adaptation of a non-parametric regression spline model. Stat Med2003; 22: 2335–2363.
46.
GlynnRJRosnerB. Comparison of risk factors for the competing risks of coronary heart disease, stroke, and venous thromboembolism. Am J Epidemiol2005; 162: 975–982.
47.
LauBColeSRGangeSJ. Competing risk regression models for epidemiologic data. Am J Epidemiol2009; 170: 244–256.
48.
GasparriniA. Modeling exposure-lag-response associations with distributed lag non-linear models. Stat Med2014; 33: 881–899.
49.
MacKenzieTAbrahamowiczM. Marginal and hazard ratio specific random data generation: applications to semi-parametric bootstrapping. Stat Comput2002; 12: 245–252.
50.
SylvestreMPAbrahamowiczM. Comparison of algorithms to generate event times conditional on time-dependent covariates. Stat Med2008; 27: 2618–2634.
51.
CoxDR. Regression models and life tables. J R Stat Soc B1972; 34: 187–220.
MahmudMAbrahamowiczMLeffondréKet al.Selecting the optimal transformation of a continuous covariate in Cox’s regression: implications for hypothesis testing. Commun Stat2006; 35: 27–45.
54.
Wright JM and Musini VM. First-line drugs for hypertension. Cochrane Database Syst Rev 2009; (3): CD001841.
55.
RayWA. Evaluating medication effects outside of clinical trials: new-user designs. Am J Epidemiol2003; 158: 915–920.
56.
World Health Organization. Guidelines for ATC classification and DDD assignment, Oslo: WHO Collaborating Centre for Drug Statistics Methodology, 2016.
57.
GreevyRAJrHuizingaMMRoumieCLet al.Comparisons of persistence and durability among three oral antidiabetic therapies using electronic prescription-fill data: the impact of adherence requirements and stockpiling. Clin Pharmacol Ther2011; 90: 813–819.
58.
SchumacherMOhnebergK. Competing risk bias was common in a prominent medical journal. J Clin Epidemiol2016; 80: 135–136.
59.
PazianasMAbrahamsenBWangYet al.Incidence of fractures of the femur, including subtrochanteric, up to 8 years since initiation of oral bisphosphonate therapy: a register-based cohort study using the US MarketScan claims databases. Osteoporos Int2012; 23: 2873–2884.
60.
LamSHarrisDRocqueBGet al.Pediatric endoscopic third ventriculostomy: a population-based study. J Neurosurg2014; 14: 455–464.
61.
Lee TA and Pickard AS. Exposure definition and measurement. In: Velentgas P, Dreyer NA, Nourjah P, et al. (eds) Developing a protocol for observational comparative effectiveness research: a user’s guide. AHRQ Publication No. 12(13)-EHC099. Rockville, MD: Agency for Healthcare Research and Quality, January 2013: Chap. 4, pp.45–58.
62.
CookJRStefanskiLA. Simulation-extrapolation estimation in parametric measurement error models. J Am Stat Assoc1994; 89: 1314–1328.
63.
BrookhartMAWangPSSolomonDHet al.Evaluating short-term drug effects using a physician-specific prescribing preference as an instrumental variable. Epidemiology2006; 17: 268–275.
64.
MacKenzieTATostesonTDMordenNEet al.Using instrumental variables to estimate a Cox's proportional hazards regression subject to additive confounding. Health Serv Outcomes Res Methodol2014; 14: 54–68.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
