Fo Fore reca cast sting and and De Deflecting ecting the Opi the - - PowerPoint PPT Presentation

▶
fo fore reca cast sting and and de deflecting ecting the
SMART_READER_LITE
LIVE PREVIEW

Fo Fore reca cast sting and and De Deflecting ecting the Opi the - - PowerPoint PPT Presentation

Aug 13, 2022 133 likes •872 views

Fo Fore reca cast sting and and De Deflecting ecting the Opi the Opioid Epidem Epidemic ic Cur Curve Donald S. Burke MD, Dean, Graduate School of Public Health, University of Pittsburgh 19 July 2017 Allegheny County Overdose Prevention

slide-1
SLIDE 1

Fo Fore reca cast sting and and De Deflecting ecting the the Opi Opioid Epidem Epidemic ic Cur Curve

Donald S. Burke MD, Dean, Graduate School of Public Health, University of Pittsburgh 19 July 2017 Allegheny County Overdose Prevention Coalition Conference

slide-2
SLIDE 2

Jeanine Buchanich, PhD Associate Professor, Biostatistics Hawre Jalal, MD, PhD Assistant Professor, Health Policy and Management

slide-3
SLIDE 3

Outline

  • 1. Epidemiology of overdoses
  • 2. Reconstructing the history of the epidemic
  • 3. Forecasting the future
  • 4. Synoptic epidemiology
  • 5. Economic burden in Western PA
  • 6. Conclusions
slide-4
SLIDE 4

Accidental Poisoning Age‐Adjusted Mortality

1981‐1983

4

slide-5
SLIDE 5

5

Accidental Poisoning Age‐Adjusted Mortality

1984‐1986

slide-6
SLIDE 6

6

Accidental Poisoning Age‐Adjusted Mortality

1987‐1989

slide-7
SLIDE 7

7

Accidental Poisoning Age‐Adjusted Mortality

1990‐1992

slide-8
SLIDE 8

8

Accidental Poisoning Age‐Adjusted Mortality

1993‐1995

slide-9
SLIDE 9

9

Accidental Poisoning Age‐Adjusted Mortality

1996‐1998

slide-10
SLIDE 10

10

Accidental Poisoning Age‐Adjusted Mortality

1999‐2001

slide-11
SLIDE 11

11

Accidental Poisoning Age‐Adjusted Mortality

2002‐2004

slide-12
SLIDE 12

12

Accidental Poisoning Age‐Adjusted Mortality

2005‐2007

slide-13
SLIDE 13

13

Accidental Poisoning Age‐Adjusted Mortality

2008‐2010

slide-14
SLIDE 14

14

Accidental Poisoning Age‐Adjusted Mortality

2011‐2013

slide-15
SLIDE 15

15

Accidental Poisoning Age‐Adjusted Mortality

Western Pennsylvania

slide-16
SLIDE 16

Top Causes of Death by Age Group in Pennsylvania

Age Group Top Cause of Death Deaths Population Rate < 1

Perinatal complications 530 142561 371.77

1‐4

Accidents due to falls, mechanical forces, drowning and suffocation 19 573343 3.31

5‐14

Motor Vehicle Accidents 21 1518107 1.38

15‐24

Motor Vehicle Accidents 285 1726198 16.51

25‐34

Overdose 567 1611833 35.18

35‐44

Overdose 502 1529401 32.82

45‐54

Ischemic Heart Disease 843 1840179 45.81

55‐64

Ischemic Heart Disease 1937 1740849 111.27

65‐74

Ischemic Heart Disease 2830 1110501 254.84

75‐84

Ischemic Heart Disease 4774 651434 732.84

85+

Ischemic Heart Disease 8037 329395 2439.93

slide-17
SLIDE 17

Male Accidental Poisoning Mortality Rates

17

25‐34 35‐44 15‐24 65+ 55‐64 45‐54

slide-18
SLIDE 18

18

Female Accidental Poisoning Mortality Rates

45‐54 35‐44 15‐24 65+ 55‐64 25‐34

slide-19
SLIDE 19

Counties with the Top 5 Largest PA Cities

Reading Philadelphia Pittsburgh Erie Allentown

19

slide-20
SLIDE 20

Rates of drug overdose deaths by state, US 2015

The Epidemic of Drug Overdose Deaths

slide-21
SLIDE 21

5 10 15 20 25 30 35 40 1970 1980 1990 2000 2010 2020

APPALACHIA CLUSTER

WV OH KY PA TN 5 10 15 20 25 30 35 40 1970 1980 1990 2000 2010 2020

NEW ENGLAND CLUSTER

NH RI MA CT 5 10 15 20 25 30 35 40 1970 1980 1990 2000 2010 2020

USA MEAN

5 10 15 20 25 30 35 40 1970 1980 1990 2000 2010 2020

REST OF THE USA

slide-22
SLIDE 22

Types of Drugs Found in the System in Fatal Overdoses Allegheny County, Pennsylvania

Heroin

Fentanyl

Number of Fatal Overdoses Ethanol Cocaine

Alprazolam

Overdose Free PA

Karl Williams, Allegheny County Medical Examiner Jan Pringle, University of Pittsburgh, School of Pharmacy

slide-23
SLIDE 23

D

Methamphetamine Cocaine Cannabis Heroin

Drugs Identified by Laboratories in Selected US Cities, 2014

National Forensic Laboratory Information System, DEA

http://www.deadiversion.usdoj.gov/nflis/NFLIS2014AR.pdf

slide-24
SLIDE 24

200 400 600 800 1,000 1,200 1,400 2,000 4,000 6,000 8,000 10,000 12,000 14,000

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Number of Users (Thousands)

Persons Aged 12 or Older

Non‐medical users of pain relievers Heroin users

Source: SAMHSA, 2014 National Survey on Drug Use and Health (September 2015).

9/2014

Heroin Use and Non-Medical Use of Pain Relievers in the Past Year among Persons Aged 12 or Older: 2002-2014

= Significant change from 2014

slide-25
SLIDE 25

Substance Dependence or Abuse for Specific Substances in the Past Year among Persons Aged 12 or Older: 2014

200 400 600 800 600 1,200 1,800 2,400 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Numbers in thousands of users

HEROIN

NONMEDICAL USE OF PAIN RELIEVERS

Source: SAMHSA, 2014 National Survey on Drug Use and Health (September 2015).

slide-26
SLIDE 26

SAMHSA 2014 National Survey on Drug Use and Health Current prevalence in the USA Non‐medical use of pain relievers 10 million Dependent 2 million Heroin users 1 million Dependent 0.6 million

DSB summary

slide-27
SLIDE 27

Reconstructing the History

  • f the Opioid Epidemic
slide-28
SLIDE 28

How did the Opioid Epidemic happen? 3 threads

Advocacy for increased treatment of chronic pain New higher dose drug formulations aggressively marketed Cheap heroin supply introduced

N Engl J Med 1980 Jan10;302(2):123. Addiction rare in patients treated with narcotics. Porter J, Jick H

Xalisco, Mexico

http://www.pennlive.com/opinion/2015/11/the_doctors_mistake_‐_telling.html http://www.salem‐news.com/articles/june012011/holder‐purdue‐ms.php http://allthingswildlyconsidered.blogspot.com/2011/05/is‐it‐pill‐mill‐how‐do‐you‐know.html http://substencoatrib.blog.com/2013/12/14/what‐does‐black‐tar‐heroin‐smell‐like/

slide-29
SLIDE 29

N Engl J Med 1980 Jan 10; 302(2): 123.

slide-30
SLIDE 30

Source: IMS Health By Calendar Quarter

slide-31
SLIDE 31

‐> Q4 2010 ‐>

http://www.wsj.com/articles/SB10001424127887323798104578453210799151732

slide-32
SLIDE 32

LaRochelle et al JAMA Intern Med. 2015;175(6):978-987.

slide-33
SLIDE 33

Heroin OxyContin

Dart et al, NEJM, 15 Jan 2015

slide-34
SLIDE 34

Sam Quinones discussed his book at Pitt Public Health on 10 November 20015

One Book, One Communit One Book, One Community

slide-35
SLIDE 35

Forecasting the Future

slide-36
SLIDE 36

Science, 4 November 2016

slide-37
SLIDE 37
  • A 50+ year repository of mortality and population data

maintained at Pitt Public Health – CDC’s National Center for Health Statistics (NCHS) – US Census Bureau

  • 120 million + records of every individual death in the USA
  • Up‐to‐date with available geographically‐specific NCHS

mortality data

  • Age, sex, race, county, ICD cause, for every death

MOIRA DATABASE (MOrtality Information, Research, and Analysis)

Jeanine Buchanich and Gary Marsh

slide-38
SLIDE 38

5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

OVERDOSE DEATHS / USA

deaths per year

slide-39
SLIDE 39

5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000

1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

OVERDOSE DEATHS / USA

deaths per year

slide-40
SLIDE 40

3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

USA Deaths Due to Accidental Poisoning, 1979‐2015

Re‐plotted on a logarithmic scale

Log 10 deaths / year

slide-41
SLIDE 41

3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

USA Deaths Due to Accidental Poisoning, 1979‐2015 Data fit to log‐linear model Log 10 deaths / year

  • Good fit for 37 years
  • Average increase per year = 9%
  • Doubling time = 8 years
slide-42
SLIDE 42

What are the forces and mechanisms behind this inexorable growth in overdose mortality?

slide-43
SLIDE 43

http://www.2020techblog.com/2016/03/at‐intel‐moores‐law‐unstables‐as‐intel.html

Moore’s Law: The number of transistors in a dense integrated circuit doubles approximately every two years

slide-44
SLIDE 44

Regardless of the mechanisms involved, we can use these historical data to forecast probable future epidemic severity

slide-45
SLIDE 45

3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

USA Deaths/Year Due to Accidental Poisoning, 1979‐2015 Extrapolation to 2016‐2020 Log 10 deaths / year

slide-46
SLIDE 46

3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

USA Deaths/Year Due to Accidental Poisoning, 1979‐2015 Forecast of number of deaths, 2016‐2020

2016 50,199 2017 54,689 2018 59,579 2019 64,908 2020 70,713 300,089

Log 10 deaths / year

slide-47
SLIDE 47

3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

USA Deaths Due to Accidental Poisoning, 1979‐2015 Forecast of number of deaths, 2016‐2020 Log 10 deaths / year Could the curve instead begin to bend downward?

slide-48
SLIDE 48

Epidemic Pattern Pennsylvania, by year

slide-49
SLIDE 49

500 1,000 1,500 2,000 2,500 3,000 3,500 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 OVERDOSE DEATHS / PENNSYLVANIA

deaths per year

slide-50
SLIDE 50

R² = 0.967 500 1,000 1,500 2,000 2,500 3,000 3,500 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 OVERDOSE DEATHS / PENNSYLVANIA

deaths per year

Pennsylvania shows the same exponential growth pattern as the entire nation

slide-51
SLIDE 51

0.5 1 1.5 2 2.5 3 3.5 4 1970 1980 1990 2000 2010 2020 2030

TX

0.5 1 1.5 2 2.5 3 3.5 1970 1980 1990 2000 2010 2020 2030

VA

0.5 1 1.5 2 2.5 3 3.5 4 1970 1980 1990 2000 2010 2020 2030

PA

0.5 1 1.5 2 2.5 3 3.5 1970 1980 1990 2000 2010 2020 2030

MO

0.5 1 1.5 2 2.5 3 3.5 4 1970 1980 1990 2000 2010 2020 2030

IL

0.5 1 1.5 2 2.5 3 3.5 4 1970 1980 1990 2000 2010 2020 2030

TN

slide-52
SLIDE 52

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 TX VA PA MO IL TN NC GA AZ WA CO OH SC FL KY AL OR NV LA OK ID VT MT NH DE AK MN WY WI MI HI ME IN KS CA MS CT UT AR NE NJ IA WV NY NM ND SD DC MA MD RI

R Squared State Log Linear Fit 1979‐2015

slide-53
SLIDE 53

Forecasts of Accidental Poisoning Deaths by State Extrapolations from 1979‐2015

State Cases (2016‐2020) AK 663 AL 4484 AR 1813 AZ 10033 CA 23242 CO 6201 CT 5476 DC 532 DE 1207 FL 32619 GA 9629 HI 1260 IA 1318 ID 1130 IL 11868 IN 4742 KS 1687 KY 8948 LA 6467 MA 3412 MD 614 ME 1568 MI 8113 MN 3320 MO 8754 State Cases (2016‐2020) MS 2198 MT 834 NC 10374 ND 169 NE 537 NH 1651 NJ 7425 NM 7249 NV 4451 NY 13009 OH 15631 OK 8483 OR 3268 PA 17488 RI 593 SC 5377 SD 222 TN 10776 TX 23685 UT 2567 VA 6985 VT 524 WA 8098 WI 5025 WV 3609 WY 596

slide-54
SLIDE 54

0.5 1 1.5 2 2.5 3 1970 1980 1990 2000 2010 2020 2030

Orange County, FL

0.5 1 1.5 2 2.5 1970 1980 1990 2000 2010 2020 2030

  • St. Louis County, MO

1 2 3 4 1970 1980 1990 2000 2010 2020 2030

Harris County, TX

1 2 3 4 1970 1980 1990 2000 2010 2020 2030

Maricopa County, AZ

0.5 1 1.5 2 2.5 3 1970 1980 1990 2000 2010 2020 2030

Allegheny County, PA

0.5 1 1.5 2 2.5 1970 1980 1990 2000 2010 2020 2030

Salt Lake County, UT

slide-55
SLIDE 55

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Allegheny County, PA Maricopa County, AZ Harris County, TX

  • St. Louis County, MO

Orange County, FL Salt Lake County, UT King County, WA Franklin County, OH Cuyahoga County, OH Hennepin County, MN Palm Beach County, FL Sacramento County, CA Broward County, FL Dallas County, TX Travis County, TX Riverside County, CA Pima County, AZ Wayne County, MI Clark County, NV Hillsborough County, FL Bexar County, TX Tarrant County, TX Mecklenburg County, NC Cook County, IL Miami‐Dade County, FL Fairfax County, VA Montgomery County, MD Contra Costa County, CA Suffolk County, NY Alameda County, CA Philadelphia County, PA Middlesex County, MA San Diego County, CA Santa Clara County, CA Queens County, NY Kings County, NY Bronx County, NY Nassau County, NY Orange County, CA Los Angeles County, CA New York County, NY Oakland County, MI San Bernardino County, CA

R Squared County Log Linear Fit 1979‐2015

slide-56
SLIDE 56

Forecasts of Accidental Poisoning Deaths by County Extrapolations from 1979‐2015

County Cases (2016‐2020) Alameda County, CA 905 Allegheny County, PA 1872 Bexar County, TX 2005 Bronx County, NY 2347 Broward County, FL 1884 Clark County, NV 3522 Contra Costa County, CA 743 Cook County, IL 5262 Cuyahoga County, OH 1517 Dallas County, TX 2065 Fairfax County, VA 272 Franklin County, OH 2582 Harris County, TX 3313 Hennepin County, MN 567 Hillsborough County, FL 2203 King County, WA 2164 Kings County, NY 3264 Los Angeles County, CA 2797 Maricopa County, AZ 8463 Mecklenburg County, NC 844 Miami‐Dade County, FL 766 County Cases (2016‐2020) Middlesex County, MA 1119 Montgomery County, MD 131 Nassau County, NY 741 New York County, NY 1391 Oakland County, MI 157 Orange County, CA 1285 Orange County, FL 1456 Palm Beach County, FL 2563 Philadelphia County, PA 3863 Pima County, AZ 2593 Queens County, NY 1580 Riverside County, CA 2002 Sacramento County, CA 1858 Salt Lake County, UT 963 San Bernardino County, CA 407 San Diego County, CA 1939 Santa Clara County, CA 585

  • St. Louis County, MO

1054 Suffolk County, NY 1061 Tarrant County, TX 1449 Travis County, TX 1203

slide-57
SLIDE 57

“Synoptic Epidemiology” Visual analysis of patterns in big epidemiological data sets

slide-58
SLIDE 58
slide-59
SLIDE 59
slide-60
SLIDE 60

Epidemic Growth in Different Age Groups

“Heat Map” : Color intensity reflects epidemic intensity in a given age group at a given time

slide-61
SLIDE 61

Epidemic Growth in Different Age Groups

slide-62
SLIDE 62

> by Drug > and by Demography (gender, race, urban/rural) Epidemic Growth in Different Age Groups

slide-63
SLIDE 63

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

Total

64 heat maps

slide-64
SLIDE 64

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

slide-65
SLIDE 65

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

slide-66
SLIDE 66

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

slide-67
SLIDE 67

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

slide-68
SLIDE 68

All drugs T 40.1 Heroin T 40.2 Prescription T 40.3 Methadone T 40.4 Synthetic T40.5 Cocaine T 40.6 Unspecified Narcotic T 40.7 Unspecified Drug

slide-69
SLIDE 69

A Paradox The trajectory of the overall epidemic of

  • verdose deaths is continuous, smooth, and

inexorable But the overall epidemic itself is a composite of multiple sub‐epidemic patterns which vary substantially according to

  • Location
  • Age
  • Gender
  • Race
  • Urbanicity
  • Drug used
slide-70
SLIDE 70
slide-71
SLIDE 71

Hawre Jalal and Mark Roberts

Data from the National Survey on Drug Use and Health and a Mathematical Model Fit to the Data

0.3% 0.2% 0.1% 6 % 4 % 2 % 3 % 2 % 1 % 2 % 1 % 15% 10% 5%

slide-72
SLIDE 72

Conclusions The epidemic of overdose deaths in the USA has been growing, inexorably and exponentially, for at least 37 years This overall pattern of exponential epidemic growth is likely to continue into the future, unless effective control measures are introduced The overall USA overdose epidemic is composed of several sub‐epidemics which vary according to year, location, drugs used, age, gender, race, and urbanicity. The epidemic in Western Pennsylvania is 2‐3 fold more severe than the national average Effective epidemic control planning should be targeted to credible local epidemic forecasts. This will require systematic collection, analysis, and modeling of local data at the state, county, and community levels

slide-73
SLIDE 73

Thank you for Thank you for your attention your attention

https://sites.google.com/site/mdpafganistan/project‐definition