Development and evaluation of a framework integrating cohort effects to predict long-term effects of interventions at the population-level

Principal researcher

Maarten J. Bijlsma

Type of research

PhD research

Supervisors

Professor Eelko Hak (FWN, PE2), Professor Stijn Vansteelandt (UGent, Statistics) and Dr. Fanny Janssen (FRW, PRC)

Collaboration

Unit Pharmacoepidemiology and Pharmacoeconomics (PE2) of the Department of Pharmacy, University of Groningen.

Summary

The population of developed countries is ageing. Healthy ageing is a top priority because health and quality of life are interrelated. Pharmaceutical interventions play a strong role in maintaining good health and thereby quality of life. However, controversy remains over the nationwide health effects of large-scale introduction of preventive pharmaceutical interventions. Accurate long term predictions of health effects are decisive in the choice for implementing certain interventions. An age-period-cohort (APC) framework integrating birth cohort effects (e.g. difference in adherence, changed guidelines) may improve the validity of predictions, but fundamental studies are lacking. We aim to build such a framework by combining methodology from the fields of demography, (pharmaco)epidemiology and statistics.. We observed cohort effects to be essential to validly assess trends in drug use that were masked by standard analyses. We propose to empirically compare and evaluate the conventional framework and the framework integrating cohort effects to accurately predict long-term effects of interventions at the population level. Within the new framework we compare the classical age-period-cohort methods used in epidemiology and demography with advanced statistical methods (e.g. recently developed techniques employing mixed modelling and Bayesian inference). To support healthy ageing intervention programs, accurate techniques for estimating their effectiveness are required.

Publications

Bijlsma, M.J., Daniel, R.M., Janssen, F. and B.L. De Stavola (2017) An assessment and extension of the mechanism-based approach to the identification of age-period-cohort models. Demography 54(2): 721-743. doi:10.1007/s13524-017-0562-6.

Bijlsma, M.J., Vansteelandt, S., Janssen, F. and E. Hak (2016) The effect of adherence to statin therapy on cardiovascular mortality: quantification of unmeasured bias using falsification end-points . BMC Public Health 16:303. doi: 10.1186/s12889-016-2986-0.

Bijlsma, M.J., Janssen, F. and E. Hak (2016) Estimating time-varying drug adherence using electronic records: extending the proportion of days covered (PDC) method. Pharmacoepidemiology & Drug Safety 25(3): 325-332. doi: 10.1002/pds.3935.

Bijlsma, M.J., Janssen, F., Bos, H.J., Kamphuizen, P.W., Vansteelandt, S. and E. Hak (2015) Association between statin use and cardiovascular mortality at the population level in the Netherlands 1994-2010. Epidemiology 26(6): 802-805. doi: 10.1097/EDE.0000000000000370.

Bijlsma, M.J., Janssen, F., Lub, R., Bos, H.J., De Vries, F.M., Vansteelandt, S. and E. Hak (2015) Birth cohort appeared to confound effect estimates of guideline changes on statin utilization. Journal of Clinical Epidemiology 68: 334-340 . doi: 10.1016/j.jclinepi.2014.10.010 .

Bijlsma, M.J., Janssen, F., Bos, H.J., Kamphuisen, P.W., Vansteelandt, S. and E. Hak (2015) Associatie tussen gebruik van statine en cardiovasculaire sterfte in Nederland 1994-2010: een ecologische studie. Medisch Farmaceutische Mededelingen 4: 23-25.

Bijlsma, M.J., Hak, E., Bos, J., De Jong-van den Berg, L.T.W. and F. Janssen (2013) Assessing the effect of a guideline change on drug use prevalence by including the birth cohort dimension: the case of benzodiazepines. Pharmacoepidemiology & Drug Safety 22: 933-941. doi: 10.1002/pds.3466.

Bijlsma, M.J., Hak, E., Bos, J.H., de Jong-van den Berg, L.T. and F. Janssen (2012) Inclusion of the birth cohort dimension improved description and explanation of trends in statin use. Journal of Clinical Epidemiology 65(10): 1052 - 1060. doi: 10.1016/j.jclinepi.2012.05.009 .

Bijlsma, M.J., Bos, H.J., De Jong-van den Berg, L.T.W. and F. Janssen (2011) Periodepatronen, leeftijdpatronen en cohortpatronen in de prevalentie van benzodiazepinegebruik en statinegebruik: de invloed van veranderde richtlijnen. Pharmaceutisch Weekblad Wetenschappelijk Platform 5(10): 169-173.

Vidra N., Bijlsma, MJ. and F. Janssen (2016) Past trends of obesity-attributable mortality in The Netherlands; An application of age-period -cohort analysis. Clinical Nutrition ESPEN 13: e66. doi: 10.1016/j.clnesp.2016.03.048. (abstract)