Royal College of Surgeons in Ireland Coláiste Ríoga na Máinleá in Éirinn

1. Royal College of Surgeons in IrelandColáiste Ríoga na Máinleá in Éirinn Enhancing rational and safe prescribing in primary care Tom Fahey Professor of General Practice, RCSI Medical School & Principal Investigator, HRB Centre for Primary Care Research

2. Overview • Background • Potentially inappropriate prescribing (PIP) • OPTI-SCRIPT • Development of intervention • Results • Summary

3. Overview • Background • Potentially inappropriate prescribing (PIP) • OPTI-SCRIPT • Development of intervention • Results • Summary

4. Background • Prescribing is a challenging and complex process • Appropriate prescribing • Potentially inappropriate prescribing (PIP) • Overprescribing, underprescribingand misprescribing • Factors that contribute to PIP

5. An overview of prescribing indicators

6. Overview cont’d • Following a ‘systematic literature search’, identified 46 different tools • English and German publications only • 36 named older people as target patients • 10 did not specify target age group • Various settings • Consensus methods used in development of 19 tools • Over-, under- and mis-prescribing

7. No perfect set of indicators • The ideal set of indicators- • Cover all aspects of appropriateness • Be developed using evidence-based methods • Show significant correlation between degree of appropriateness and clinical outcomes • Be applicable not only in research but in daily health care practice Kaufmann et al, 2013

8. What contributes to PIP? • Multimorbidity • “Presence of two or more long-term conditions” • 64.9% of people aged 65-84years [1] • 30.4% of people aged 45-64 years [1] • Polypharmacy • “the ingestion of four or more medications”

9. Prevalence of PIP • PIP is prevalent in the older population (> 70 years) • Republic of Ireland 36% • Northern Ireland 34% • United Kingdom 29%

10. The prevalence of the most common STOPP/START PIP indicators across three regions

11. Overview • Background • Potentially inappropriate prescribing (PIP) • OPTI-SCRIPT • Development of intervention • Results • Summary

12. OPTI-SCRIPT study development

13. Study design & methodology – cluster RCT • GPs inclusion criteria: • Based in greater Dublin area • 80+ patients aged over 70 • Patients inclusion criteria: • Aged 70+ • Had PIP as per study list • Recruited and baseline data collection prior to minimisation

14. Study overview PCRS – National Contemporaneous Control - Observational comparison to national prescribing data (376,858 patients, 2,000+ practices)

15. OPTI-SCRIPT website

17. OPTI-SCRIPT RCT results • Participants • 21 GP practices (32% cluster response rate) • 196 patients (37% response rate) • Minimisation

18. Study design & methodology – cluster RCT • Primary outcome measure: • Proportion of patients with no PIP • Mean PIP per group • Data collection baseline & immediate post intervention • Between group differences: • Random effects logistic regression • Cluster mean • Random effects poisson regression • Process evaluation

19. Outcome – Proportion with no PIP Adjusted odds ratio = 3.06 (95% CI 1.4,6.5; P=0.004)* *adjusted for gender, age, baseline PIP, number repeat medications, GP practice size

20. National contemporaneous control – PCRS • Intervention period, Sep 2012 – August 2013 prevalence of 38% • Odds of having no PIP in OPTI-SCRIPT intervention compared to odds of having no PIP in the national PCRS cohort

21. Process evaluation – main findings • Participants positive about study • Barriers identified: GP time, communication, reimbursement • Revealed intervention not delivered as expected: • Patient information leaflets not used at all • 1 intervention practice did not complete reviews • 2 Intervention practices conducted reviews without patients • 2 control practices did alter patient medication

22. Future work • National trial of OPTI-SCRIPT • Lessons learned? • Computerise PIP identification • Focus on top 10 PIP • Embedded in practice software • Practice incentives – reimbursement • Economic evaluation

23. Overview • Background • Potentially inappropriate prescribing (PIP) • OPTI-SCRIPT • Development of intervention • Results • Summary

24. Summary • Prevalence of PIP high in Ireland & UK • Developed web-based intervention to target PIP in primary care • Process evaluation gave insight into intervention delivery and barriers • Further implementation of decision support to improve quality & safety are planned

25. Acknowledgements This research is funded by the HRB Centre for Primary Care Research and the HRB PhD Scholars programme in Health Service Research Barbara Clyne, Susan Smith, Marie Bradley, Carmel Hughes, Janine Cooper, Fiona Boland, Ronan McDonnell, David Williams, Nicola Motterlini, Marie-Claire Kennedy, Daniel Clear, Frank Moriarty, Caitriona Cahir

28. Baseline characteristics

29. PIP at baseline

30. PPI • 60% of participants had a PPI • 53% of intervention, 65% of control at baseline • At follow-up the odds of not having a PPI at maximum therapeutic dose were 3 times higher in intervention than control (OR = 3.41, P = 0.006, 95% CI 1.43, 8.14)