1. Burden, Access, and Unmet Need: the mental health service landscape in Ontario Association of General Hospital Psychiatric Services Paul Kurdyak MD PhD

2. Disclosures  Salary Support from: ICES CIHR

3. Overview 1. The burden of mental illness and addictions 2. Medical Comorbidity 3. Access to psychiatrists 4. Increasing help-seeking behaviour – a CAMH natural experiment

4. 3 Opening Eyes, Opening Minds: The Ontario Burden of Mental Illness and Addictions Report

5. Burden of Mental Illness and Addictions in Ontario  A collaboration between PHO and ICES  Involved CAMH scientists  Important because: Sets a baseline for evaluating future public health or population-based interventions Has fostered relationships between mental health and public health

6. Unit of Measurement: HALY  HALY: Health-Adjusted Life Years  HALY = YLL + YERF  YLL: Years of life lost due to premature mortality  YERF: Equivalent years of healthy life lost due to disease/disability

7. Disease Categories  Mental Health Conditions Agoraphobia Bipolar disorder Major depression Panic disorder Schizophrenia Social phobia  Addictions Alcohol use disorders Cocaine use disorders Prescription opioid misuse

8. HALYs by Mental Health Condition/ Addiction

9. YLL by Mental Health Condition/ Addiction YLLs by Mental Health Condition/ Addiction

10. YERF by Mental Health Condition/ Addiction YERFs by Mental Health Condition/ Addiction

11. HALYs by Age Group 10

12. Comparison to Other BoD Studies Cancers MI&A Infectious Diseases

13. Overview 1. The burden of mental illness and addictions 2. Medical Comorbidity 3. Access to psychiatrists 4. Increasing help-seeking behaviour – a CAMH natural experiment

14. Mortality Burden Dramatically Under-estimated  Cause of death is disease-specific.  No one dies from schizophrenia  Premature mortality in schizophrenia mostly due to cardiovascular disease and risk factors  Access to medical care is very poor

15. MaleFemale Crude Rate Ratio Age Adjusted RR (95% CI) Crude Rate Ratio Age Adjusted RR (95% CI) SCZ1.842.51 (2.43, 2.60)2.642.34 (2.26, 2.42) BPD1.802.00 (1.95, 2.05)1.641.89 (1.85, 1.94) All Cause Mortality: SCZ and BPD (2006- 2010)

16. Schizophrenia Outcomes Following AMI 15 89,825 AMI Subjects 1087 Allocated to Schizophrenia 88,738 Allocated to No Schizophrenia 842 with Schizophrenia 70,826 without Schizophrenia Excluded: 8 – Missing Data 81 – Not Incident AMI 156 – Death before Discharge Excluded: 8 – Missing Data 81 – Not Incident AMI 156 – Death before Discharge Excluded: 394 – Missing Data 7628 – Not Incident AMI 9890 – Death before Discharge Excluded: 394 – Missing Data 7628 – Not Incident AMI 9890 – Death before Discharge Mortality Outcome 809 with Schizophrenia 69,102 without Schizophrenia Excluded: 33 – Death within 30 days of discharge Excluded: 33 – Death within 30 days of discharge Excluded: 1724 - Death within 30 days of discharge Excluded: 1724 - Death within 30 days of discharge Process of Care Outcome

17. Mortality 16 Unadjusted Adjusted AOR 1.56, 95% CI 1.08-2.23; p=0.02

18. Cardiac Procedures 17 UnadjustedAdjusted AOR 0.48, 95% CI 0.40-0.56; p<0.001

19. Cardiologist Visits 18 UnadjustedAdjusted AOR 0.53, 95% CI 0.43-0.65; p<0.001

20. Overview 1. The burden of mental illness and addictions 2. Medical Comorbidity 3. Access to psychiatrists 4. Increasing help-seeking behaviour – a CAMH natural experiment

21. Ability to Access Psychiatrists  Primary care physician surveys from multiple jurisdictions - psychiatrists most difficult specialists to access  NPS survey 2007 - from 2004 to 2007, ability to accept urgent referral (< 1 week) increased from 44% to 49%  Other specialties increased from 60% (2004) to 80% (2007)  2010 survey – 35% primary care physicans rated access to psychiatrists as poor (vs. 4% of GIM and 2% for pediatricians) 20

22. 21 160 Unavailable (70%) 297 Psychiatrists 230 Contacted

23. 22 297 Psychiatrists 230 Contacted 64 (27%) Need to review referral information and no wait-time estimate 160 Unavailable (70%)

24. 23 297 Psychiatrists 230 Contacted 6 (3%) offered immediate appointments (wait times 4-55 days) 64 (27%) Need to review referral information and no wait-time estimate 160 Unavailable (70%)

25. Ontario Psychiatrist Supply  Toronto and Ottawa have 2-4 times more psychiatrists per capita than other regions in Ontario. 24

26. What Are Psychiatrists Doing?  There are large differences between psychiatrist supply across different regions  Toronto and Ottawa have large supplies per capita  The rest of the province hovers around 10 psychiatrists/100,000  If there are so many psychiatrists (and so many more in Toronto and Ottawa), why are they the most difficult to access? 25

27. Mean # Unique Patients and # New Patients per Year  Low supply area psychiatrists see twice as many patients and twice as many new patients/year 26

28. Psychiatrists vs Patients in Toronto 25% of psychiatrists see 6% of outpatients 27

29. Patient Income Across Visit Categories - Toronto  Almost half of patients seen >16 times/year are in the top income quintile 28

30. Summary  Psychiatrists in high supply areas see fewer patients, fewer new patients and see these fewer patients more frequently and for longer per visit  In high supply areas, as visit frequency increases, patient SES increases  The increased psychiatrist supply does not translate into better follow-up post-hospitalization  Access to psychiatrists does not improve with increased per capita supply 29

31. Follow-up 30 days Post-Hospitalization 30

32. Readmission 31-60 days Post-Hospitalization 31

33. Summary 1. The burden of mental illness and addictions 2. Medical Comorbidity 3. Access to psychiatrists 4. Increasing help-seeking behaviour – a CAMH natural experiment 32

34. Mental Illness and Addiction Treatment Rates  Two thirds of people with depression do not seek help  Up to 90% of people with addictions do not seek treatment  Very little evidence on increasing treatment-seeking behaviours to address burden of mental illness and addiction

35. The CAMH Campaign

36. A Natural Experiment  The campaign is the only intervention that occurred in March 2010 (nothing else changed that could explain changes in visit volumes)  Permits an evaluation of the campaign using quasi- experimental methods  ED volumes AND Gen Psych. Assessment Clinic volumes – direct-to-consumer marketing vs. service provider marketing

37. Methods  All patients who presented to the ED (N=29,069) and the Gen Psych. Assessment Clinic (N=8326) from April 1, 2006 to December 31, 2011.  Grouped monthly  Pre-campaign – April 1, 2006 to March 31, 2010  Post-campaign – April 1, 2010 to December 31, 2011  Also used regional-level data for system-level analyses (preliminary)

38. Statistical Analysis  Time series analysis methods used to model the data series and test for an effect of the campaign.  Geographic Information Systems (GIS) using patient postal code for mapping patient distance from ED.

39. ED Volumes

40. General Psychiatry Assessment Clinic Volumes

41. ED Volumes: % new to CAMH and Region

42. Pre-Campaign Map

43. Post-Campaign Map

44. Maps Side by Side

45. Limitations  Just starting system context Don’t know if we are duplicating services Preliminarily – campaign increased volume in all categories: previous CAMH ED visit, new to CAMH, and new to region

46. Main Findings  Addressing stigma increases help-seeking and referral behaviour  Can have a significant impact on volumes  Low treatment rates can be addressed using marketing strategies addressing stigma AND highlighting service availability

47. Summary  Huge burden of mental illness and addictions in Ontario  High supply of psychiatrists in Toronto and incentivization are perpetuating poor access in the face of very high psychiatrist supply  Access to care at high times of need (post- hospitalization) is poor  CAMH campaign suggests there is a large unmet need “market” that is currently not being served