Pdmp 4 kreiner small

PDMP  Track:  Prescrip/on  Behavior  Surveillance  System:  The  Value  and  

Applica/ons  of  De-­‐iden/fied  PDMP  Data  in  Public  Health  Surveillance  

Presenters:  

Dr.  Peter  W.  Kreiner,  PhD,  PDMP  Principal  Inves7gator,  Prescrip7on  Drug  Monitoring  Program  Center  of  Excellence,  Brandeis  University  

Mike  Small,  DOJ  Administrator,  Law  Enforcement  Support  Program,    CA  Department  of  Jus7ce    

Moderator:  John  L.  Eadie,  Director,  Prescrip7on  Drug  Monitoring  Program  Center  of  Excellence,  Brandeis  University  

Disclosures  

•  Dr.  Peter  W.  Kreiner  has  disclosed  no  relevant,  real  or  apparent  personal  or  professional  financial  rela8onships.  

•  Mike  Small  has  disclosed  no  relevant,  real  or  apparent  personal  or  professional  financial  rela8onships.  

Learning  Objec/ves  

1.  Explain  the  value  and  applica8ons  of  de-­‐iden8fied  PDMP  data.    

2.  List  addi8onal  resources  for  op8mum  data  encryp8on  within  the  PDMP  program.    

3.  State  the  benefits  of  coordina8ng  interstate  data  sharing  and  the  barriers  that  currently  exist  to  limit  these  agreements.  

The  Prescrip/on  Behavior  Surveillance  System:  Applica/ons  of  De-­‐iden/fied  PDMP  Data  in  

Public  Health  Surveillance  

Rx  Abuse  Summit  

April  23,  2014  

Peter  Kreiner,  Ph.D.  

PDMP  Center  of  Excellence  at  Brandeis  University  

The  Prescrip/on  Behavior  Surveillance  System  (PBSS)  

A  longitudinal,  mul8-­‐state  database  of  de-­‐iden8fied  PDMP  data,  to  serve  as:  

1.  An  early  warning  surveillance  tool  2.  An  evalua8on  tool,  in  rela8on  to  state  and  local  

policies  and  ini8a8ves,  such  as  prescriber  educa8onal  ini8a8ves  

PBSS  Con/nued  

Began  in  FY2012  with  support  from  CDC  and  FDA,  administered  through  BJA  Guided  by  Oversight  CommiVee:  –  Federal  partners:  CDC,  FDA,  BJA,  SAMHSA  –  State  partners  to  date:  CA,  DE,  FL,  ID,  ME,  OH    

–  Addi8onal  state  partners  in  process  (IN,  KY,  WA)  –  Adjunct  state  partners  (MA,  OK,  TN)  –  unable  to  share  data  but  willing  to  provide  PBSS  surveillance  measures  

–  No  release  of  data  or  findings  without  Oversight  CommiVee  approval  

PBSS  Con/nued  

De-­‐iden8fied  data  from  each  par8cipa8ng  state  –  Data  use  agreements  tailored  to  each  state’s  laws  and  requirements  

–  Beginning  with  2010  or  2011,  ini8al  2  –  3  years  of  data  –  Data  updated  quarterly  (including  prior  12  months)  –  Project-­‐specific  ID  #’s  for  pa8ents,  prescribers,  pharmacies  

•  Maintained  for  the  dura8on  of  the  data  –  Data  housed  in  secure  IT  environment  at  Brandeis  University  

PBSS  Con/nued    

Purpose  (1):  early  warning  public  health  surveillance  tool  –  Periodic  surveillance  report  with  40+  pa8ent,  prescriber,  and  pharmacy  measures  

–  Help  iden8fy  emerging  trends,  changes  in  prescrip8on  paVerns,  indicators  of  risk  associated  with  new  drugs/new  drug  formula8ons  

–  Data  dic8onary  and  technical  documenta8on  –  Development  of  standards  in  rela8on  to  data  quality  (error  rates,  missing  fields)  and  pa8ent  linking  (determining  which  prescrip8on  records  belong  to  the  same  pa8ent)  

–  Online  access  for  authorized  federal  researchers  –  Specialized  analyses  as  agreed  upon  with  each  state  partner  

PBSS  Con/nued  

Purpose  (2):  evaluate  selected  ini8a8ves  to  influence  prescriber  behavior    –  Inventory  of  state  and  local  prescriber  ini8a8ves  and  

 assessment  of  their  evidence  base  •  E.g.,  state  mandates  for  prescribers  to  register  with  and  use  the  PDMP  

–  Development  and  valida8on  of  (popula8on-­‐level)  measures  of  (in)appropriate  and  aberrant  prescribing  

–  Evalua8on  of  selected  REMS  trainings  

Surveillance  Data:  Selected  Examples  •  Opioid  prescrip8on  rates  by  age  group,  comparison  

across  states  •  Number  of  Schedule  II  –  V  opioids  per  1,000  state  residents  

(in  each  age  group)  

•  Mul8ple  provider  episode  rates  by  age  group,  comparison  across  states  •  Number  of  pa8ents  with  prescrip8ons  from  5  or  more  

prescribers  and  5  or  more  pharmacies  in  a  3-­‐month  period,  per  100,000  state  residents  

•  Mul8ple  provider  episode  rates  by  quarter,  2010  –  2013  (Ohio)  

•  Mul8ple  provider  episode  rates  by  community:  MassachuseVs  

Further  Data  Examples  

•  Propor8on  of  total  prescrip8ons  accounted  for  by  prescriber  deciles:  prescriber  10%  groupings  based  on  prescrip8on  volume  

•  Average  daily  opioid  dosage  (morphine  milligram  equivalents)  

•  Average  daily  opioid  dosage  by  prescriber  decile  (based  on  volume  of  opioid  prescrip8ons  

Further  Data  Applica8ons  

•  Community  profiles  based  on  PDMP  measures  •  Inform  local  preven8on  and  treatment  needs  assessment  •  Help  evaluate  local  interven8ons/track  community  

progress  in  addressing  prescrip8on  drug  issues  

•  Mass.  PMP/Brandeis  are  developing  community  profiles  for:    •  BJA  Data  Sharing  Pilot  grantee  (Norfolk  County  District  

AVorney’s  Office)  

•  CSAP/Bureau  of  Substance  Abuse  Services  grantees  addressing  non-­‐medical  use  of  prescrip8on  drugs  

•  BSAS  block  grant-­‐funded  grantees  addressing  opioid  abuse  

Further  Data  Applica8ons,  Con8nued  

•  Versions  of  PBSS  surveillance  reports/measures  for  PDMP  cons8tuencies:  •  Governor,  legislators,  other  state  agencies,  consumer  

groups,  etc.  

•  Epidemiological  analyses  tailed  to  state  needs  •  Examina8on  of  pa8ent,  prescriber,  and  dispenser  behavior  

over  8me:  what  earlier  paVerns  are  associated  with  (eventual)  risk  indica8ons?  

•  E.g.,  paVerns  of  risk  indicators  sugges8ve  of  collec8ve  ac8vity,  drug  rings  

•  More  detailed  examina8on  of  prescribing  paVerns  for  children  or  elderly  persons  

A  Concluding  Note  

•  Increasing  use  of  PDMP  data  highlights  a  need  for  data  quality  and  data  consistency  •  States  vary  in  measures  to  ensure  and  improve  data  quality  

•  States  vary  in  procedures  to  determine  which  prescrip8on  records  belong  to  the  same  pa8ent  

•  An  important  part  of  PBSS  is  the  development  of  methods  to  assess  data  quality  and  to  assess  the  adequacy  of  prescrip8on  linking  procedures  

•  And  to  assist  PBSS  states  in  implemen8ng  best  prac8ces  for  data  quality  and  linking  procedures,  and  processes  for  ongoing  quality  improvement  

Contact  Informa/on  

Peter  Kreiner,  Ph.D.  Principal  Inves8gator  

PDMP  Center  of  Excellence  Brandeis  University  

781-­‐736-­‐3945  [email protected]    

www.pdmpexcellence.org