january 2013 quality and safety narrative report

28
GHS Quality and Safety Report January 2013 Core Measures / Value Based Purchasing Background – The Center for Medicare and Medicaid Services (CMS) and The Joint Commission (TJC) have developed process of care measures for Acute Myocardial Infarction (AMI), Congestive Heart Failure (CHF), Community Acquired Pneumonia (CAP), and Surgical Care Improvement (SCIP) termed “Core Measures”. The term “All Care Measure” (ACM) refers to perfect care provided to a patient with a specific disease. It is the percent of patients who received all the needed core measures required for that disease state. The term “Composite” refers to the percent compliance of all possible opportunities (the total number of compliant opportunities for care divided by the total number of opportunities for care). The Composite score will always be higher than the All Care Measure Score. The measures differ slightly between CMS and TJC and are publicly reported on their respective websites (CMS) www.hospitalcompare.hhs.gov and (TJC) www.qualitycheck.org. Reported results lag 3 to 6 months behind due to the complexity and requirements of external reporting. Over the past several years, we have set an organization wide goal for the All Care Measure (ACM). Beginning in FY 2012, we are changing our organizational goal to a Value Based Purchasing (VBP) score, but will continue to report the ACM and Composite scores. The Deficit Reduction Act of 2005 directed CMS to develop a Value Based Purchasing (VBP) incentive program to begin to align Medicare payments with hospital quality performance. The Patient Care and Affordable Care Act put in place the mechanism and requirement for CMS to withhold a percentage of Medicare reimbursement and require hospitals to meet performance thresholds to earn back the withheld percentage. The amount CMS will withhold in FY 2013 is 1.0% of a facility’s CMS baseline DRG payment. This withhold will increase by 0.25% annually to 2.0% in FY 2017. Based on a hospital’s total performance score, hospitals will have their DRG payments adjusted by a factor somewhere between a loss of the entire withhold, to a gain of an amount equal to the withhold. The VBP program is budget neutral resulting in many hospitals losing money and others gaining money. To be eligible, hospitals must also continue to submit results to the Hospital Compare website. The total performance score of the VBP program is a combination of several measures. Hospitals will have two methods to gain points toward their total VBP score. For each measure, a hospital can either achieve a certain level of performance or they can obtain points for improving their scores as compared to their baseline data. CMS will count the greater of the two scores, achievement versus improvement. Because GHS has historically done very well on core measures, our opportunity for improvement is minimal and our clinical domain score will likely be determined primarily by our achievement score. CMS has established national benchmarks and thresholds for each VBP quality measure. The benchmarks represent the highest achievement levels whereas the thresholds represent the minimum achievement levels. Each of our four acute care facilities will receive their own VBP score and will each

Upload: others

Post on 12-Sep-2021

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: January 2013 Quality and Safety Narrative Report

GHS  Quality  and  Safety  Report  January  2013  

 Core  Measures  /  Value  Based  Purchasing  

 Background  –  The  Center  for  Medicare  and  Medicaid  Services  (CMS)  and  The  Joint  Commission  (TJC)  have  developed  process  of  care  measures  for  Acute  Myocardial  Infarction  (AMI),  Congestive  Heart  Failure  (CHF),  Community  Acquired  Pneumonia  (CAP),  and  Surgical  Care  Improvement  (SCIP)  termed  “Core  Measures”.      The  term  “All  Care  Measure”  (ACM)  refers  to  perfect  care  provided  to  a  patient  with  a  specific  disease.    It  is  the  percent  of  patients  who  received  all  the  needed  core  measures  required  for  that  disease  state.    The  term  “Composite”  refers  to  the  percent  compliance  of  all  possible  opportunities  (the  total  number  of  compliant  opportunities  for  care  divided  by  the  total  number  of  opportunities  for  care).    The  Composite  score  will  always  be  higher  than  the  All  Care  Measure  Score.    The  measures  differ  slightly  between  CMS  and  TJC  and  are  publicly  reported  on  their  respective  websites  (CMS)  www.hospitalcompare.hhs.gov  and  (TJC)  www.qualitycheck.org.    Reported  results  lag  3  to  6  months  behind  due  to  the  complexity  and  requirements  of  external  reporting.      Over  the  past  several  years,  we  have  set  an  organization  wide  goal  for  the  All  Care  Measure  (ACM).    Beginning  in  FY  2012,  we  are  changing  our  organizational  goal  to  a  Value  Based  Purchasing  (VBP)  score,  but  will  continue  to  report  the  ACM  and  Composite  scores.    The  Deficit  Reduction  Act  of  2005  directed  CMS  to  develop  a  Value  Based  Purchasing  (VBP)  incentive  program  to  begin  to  align  Medicare  payments  with  hospital  quality  performance.    The  Patient  Care  and  Affordable  Care  Act  put  in  place  the  mechanism  and  requirement  for  CMS  to  withhold  a  percentage  of  Medicare  reimbursement  and  require  hospitals  to  meet  performance  thresholds  to  earn  back  the  withheld  percentage.    The  amount  CMS  will  withhold  in  FY  2013  is  1.0%  of  a  facility’s  CMS  baseline  DRG  payment.    This  withhold  will  increase  by  0.25%  annually  to  2.0%  in  FY  2017.    Based  on  a  hospital’s  total  performance  score,  hospitals  will  have  their  DRG  payments  adjusted  by  a  factor  somewhere  between  a  loss  of  the  entire  withhold,  to  a  gain  of  an  amount  equal  to  the  withhold.    The  VBP  program  is  budget  neutral  resulting  in  many  hospitals  losing  money  and  others  gaining  money.    To  be  eligible,  hospitals  must  also  continue  to  submit  results  to  the  Hospital  Compare  website.        The  total  performance  score  of  the  VBP  program  is  a  combination  of  several  measures.    Hospitals  will  have  two  methods  to  gain  points  toward  their  total  VBP  score.    For  each  measure,  a  hospital  can  either  achieve  a  certain  level  of  performance  or  they  can  obtain  points  for  improving  their  scores  as  compared  to  their  baseline  data.    CMS  will  count  the  greater  of  the  two  scores,  achievement  versus  improvement.    Because  GHS  has  historically  done  very  well  on  core  measures,  our  opportunity  for  improvement  is  minimal  and  our  clinical  domain  score  will  likely  be  determined  primarily  by  our  achievement  score.    CMS  has  established  national  benchmarks  and  thresholds  for  each  VBP  quality  measure.    The  benchmarks  represent  the  highest  achievement  levels  whereas  the  thresholds  represent  the  minimum  achievement  levels.    Each  of  our  four  acute  care  facilities  will  receive  their  own  VBP  score  and  will  each  

Page 2: January 2013 Quality and Safety Narrative Report

  2  

be  susceptible  to  incentive  payments  or  penalties.    (North  Greenville  Memorial  Hospital  as  a  Long  Term  Acute  Care  Hospital  is  not  eligible  for  inclusion  under  the  current  VBP  program.)    Year  1  of  the  VBP  program  involves  quality  metrics  obtained  for  services  between  July  2011  and  March  2012.    The  payment  changes  are  applied  to  DRG  payments  beginning  with  FY  2013  in  October  2012.    During  this  period,  CMS  used  two  types  of  measures.    The  first  domain  is  called  the  Clinical  Process  of  Care  Measures  and  included  12  core  measures.    This  domain  reflected  70%  of  the  total  VBP  score.    The  second  domain  is  called  the  Patient  Experience  of  Care  Measures  and  included  8  HCAHPs  patient  satisfaction  measures.    This  domain  reflected  30%  of  the  total  VBP  score.    In  FY13,  all  GHS  hospitals  received  an  increase  to  their  DRG  payments  based  on  our  quality  results.    CMS  intends  to  add  measures  to  the  VBP  program.    In  Year  2,  CMS  added  one  additional  core  measure  to  the  Clinical  Process  of  Care  Measures  involving  the  removal  of  urinary  catheters.    They  are  weighting  this  domain  at  45%.    The  Patient  Experience  of  Care  domain  is  weighted  at  30%.    They  have  added  a  third  domain,  termed  Outcome  of  Care  and  weighted  it  at  25%.    It  is  a  measure  of  risk-­‐adjusted  mortality  rates  for  acute  myocardial  infarction,  congestive  heart  failure  and  community  acquired  pneumonia.    The  measurement  period  is  from  April  2012  through  December  2012.    Adjustments  to  DRG  payments  will  be  made  in  October  2013  at  the  start  of  FY  2014.    GHS  Goal  –  For  FY  2013,  the  GHS  quality  goal  is  the  new  measure  for  Value  Based  Purchasing.    Specifically,  it  is  the  composite  compliance  score  for  the  13  clinical  core  measures.    Historically,  our  composite  score  for  these  measures  has  been  around  98%,  which  approximates  the  75th  percentile.    Thus,  the  GHS  goal  is  set  at  98.0%  to  maintain  performance  at  this  level.    We  will  continue  to  report  the  ACM  and  Composite  scores.        GHS  Results  –    

Value  Based  Purchasing  –  The  initial  results  for  the  6  month  period  April  2012  through  September  2012  for  all  four  acute  care  facilities  exceed  our  target  of  98.0%  except  for  Greenville  Memorial.    The  GHS  overall  VBP  clinical  process  of  care  score  is  98.6%,  Greenville  Memorial’s  score  is  97.4%,  Greer  Memorial’s  is  99.0%,  Hillcrest  Memorial’s  score  is  98.4%,  and  Patewood  Memorial’s  score  is  99.7%.        ACM  /  Composite  Scores  –  From  October  2011  through  September  2012,  the  GHS  ACM  compliance  rate  is  95.9%  for  inpatient  measures,  97.9%  for  outpatient  measures,  and  96.2%  combined.    The  inpatient  composite  compliance  rate  for  this  time  period  is  98.8%.        The  Acute  Myocardial  Infarction  (AMI)  ACM  score  for  July  -­‐  September  12  is  99.0%,  while  the  composite  compliance  rate  is  99.7%  (763/765).        The  Congestive  Heart  Failure  (HF)  ACM  score  for  July  -­‐  September  12  is  99.5%,  while  the  composite  compliance  rate  is  99.8%  (400/401).        

Page 3: January 2013 Quality and Safety Narrative Report

  3  

The  Community  Acquired  Pneumonia  (CAP)  ACM  score  for  July  -­‐  September  12  is  97.7%,  while  the  composite  compliance  rate  is  98.5%  (261/265).        The  Surgical  Care  (SCIP)  ACM  score  for  July  -­‐  September  12  is  93.7%,  while  the  composite  compliance  rate  is  99.0%  (4157/4201).        The  Outpatient  Emergency  Department  (ED)  and  Surgical  All  Care  Measure  score  for  July  -­‐  September  12  is  97.6%  while  the  composite  score  is  98.7%  (517/524).      

 Specif ic   Issues  –  The  primary  opportunity  for  improvement  involves  the  removal  of  post-­‐operative  urinary  catheters  within  2  days  of  surgery  in  order  to  prevent  a  catheter  associated  urinary  tract  infection  (CAUTI).  An  innovative  build  in  our  Soarian  technology  has  led  to  the  development  of  an  alert  in  using  Computerized  Physician  Order  Entry  (CPOE)  that  is  generated  to  the  surgeon  on  post  operative  day  1  or  2.    The  electronic  alert  prompts  the  physician  to  either  discontinue  the  urinary  catheter  or  to  document  an  evidence-­‐based  reason  for  continuing  the  catheter.    The  ‘go-­‐live’  date  was  January  9,  2013.    Concurrent  monitoring  will  provide  feedback  on  the  effectiveness  of  this  latest  intervention.      

Mortal ity  Rates    

Background  –  We  assess  mortality  rates  through  four  methods.        

CMS  30  Day,  All  Cause  Mortality  Rates  for  AMI  /  CHF  /  Pneumonia  –  CMS  calculates  and  reports  30  day,  all-­‐cause  mortality  rates  for  patients  admitted  with  AMI,  CHF,  or  pneumonia  on  their  public  website  at  www.hospitalcompare.hhs.gov.      Because  they  have  complete  claims  and  eligibility  data,  they  are  able  to  identify  patients  who  die  after  being  admitted  to  any  hospital  in  the  country.    CMS  calculates  this  data  once  annually.    The  current  measures  are  for  July  2008  through  June  2011.      

 Premier  In-­‐Hospital  Mortality  Rates  –  We  assess  system,  facility,  and  DRG  business  line  level  data  of  all-­‐cause,  in-­‐hospital  mortality  throughout  GHS  utilizing  the  Premier  Clinical  Advisor  database.    A  mortality  rate  index  is  calculated  that  represents  a  risk-­‐adjusted  measure  of  the  observed  mortality  rate  divided  by  the  expected  mortality  rate.        AHRQ  Inpatient  Quality  Indicators  (IQIs)  –  The  Agency  for  Healthcare  Research  and  Quality  (AHRQ)  has  developed  the  Inpatient  Quality  Indicators  (IQIs),  which  are  a  set  of  measures  that  provide  perspective  on  hospital  quality  of  care  using  hospital  administrative  (claims)  data.    The  data  source  for  AHRQ  IQI  data  is  provided  by  CMS  on  an  annual  basis  to  all  participating  hospitals  across  the  country.    The  benchmarks  in  the  CMS  annual  report  are  derived  from  their  national  database.    At  this  time  CMS  is  scheduled  to  publicly  report  on  their  Hospital  Compare  website  only  two  of  the  AHRQ  IQI  indicators,  Hip  Fracture  Mortality  Rate  and  AAA  (Abdominal  

Page 4: January 2013 Quality and Safety Narrative Report

  4  

Aortic  Aneurysm)  Repair  Mortality  Rate.      In  this  section,  we  are  presenting  data  for  the  IQIs  that  assess  inpatient  mortality  rates  only.      

 GHS  Site-­‐Specific,  5-­‐Year  Cancer  Survival  Rates  –  Annually,  we  review  our  5-­‐year  cancer  survival  rates  for  several  specific  forms  of  cancer  as  part  of  our  cancer  care  accreditation.    The  data  are  obtained  from  our  cancer  registry  and  compared  to  the  National  Cancer  Database  (NCDB)  national  benchmarks.    The  September  2012  study  performed  by  Dawn  Blackhurst,  DrPH  assessed  the  5-­‐year  survival  of  “analytic”  cases  diagnosed  with  cancer  in  2003,  2004  and  2005.    “Analytic”  cancer  cases  are  those  who  were  diagnosed  or  received  their  first  course  of  treatment  at  GHS.    GHS  survival  rates  were  compared  to  rates  from  all  NCDB  hospitals  (n=1474  hospitals).    Rates  were  formally  compared  for  statistical  significance  using  95%  confidence  intervals.        

 GHS  Goal  –  Our  goal  is  for  our  mortality  index  or  rates  to  be  statistically  better  than  expected.    For  the  IQIs,  our  goal  is  to  have  a  rate  lower  than  the  comparative  benchmark.    GHS  Results    

CMS  30-­‐Day,  All  Cause  Mortality  Rates  for  AMI  /  CHF  /  Pneumonia  –  Our  mortality  rates  for  July  2008  through  June  2011,  reported  in  2012  are  statistically  no  different  than  the  national  average.    Note  that  as  the  population  becomes  smaller  around  a  specific  disease,  it  is  very  difficult  to  show  statistical  significance.  

 Premier  In-­‐Hospital  Mortality  Rates  –  Our  system  wide  in-­‐hospital,  all-­‐cause  mortality  rate  for  October  2011  through  September  2012  remains  constant  at  2.2%  with  a  mortality  rate  index  of  0.85.    This  is  statistically  better  than  expected  for  the  Greenville  Health  System  as  well  as  for  the  individual  facilities  of  Greenville,  Greer  and  Hillcrest.    Patewood  has  a  0.00%  mortality  rate,  but  due  to  lower  volumes,  does  not  reach  statistical  significance.    DRG  level  mortality  rate  indices  are  presented  for  Greenville  Hospital  System  as  a  whole  with  no  major  opportunities  identified.        AHRQ  Inpatient  Quality  Indicators  (IQIs)  –  For  Greenville  Memorial  we  have  the  ability  to  benchmark  AHRQ  IQI  results  with  other  UHC  teaching  hospitals.    The  IQI  mortality  rates  are  statistically  unfavorable  to  the  UHC  benchmark  for  one  of  the  Inpatient  Quality  Indicators,  craniotomy.    Several  other  IQI  results  are  statistically  higher  than  Premier  peer  rates,  but  these  comparative  hospitals  have  a  lower  acuity  of  patients  making  the  UHC  benchmark  more  comparable  for  GMH.    The  community  hospitals  have  much  lower  volumes  and  are  not  statistically  different  for  the  procedures  included  in  the  AHRQ  Inpatient  Quality  Indicators.    GHS  Site-­‐Specific,  5-­‐Year  Cancer  Survival  Rates  –  Overall  “combined-­‐stage”  GHS  5-­‐year  survival  rates  were  comparable  to  NCDB  rates  for  all  11  cancer  sites  [See  Figure  1   in  Appendix  A].    GHS  survival  rates  were  slightly  higher  than  NCDB  rates  for  9  of  11  cancer  sites  and  slightly  lower  for  2  of  the  11  sites;  however,  none  of  these  differences  were  statistically  significant.      

 

Page 5: January 2013 Quality and Safety Narrative Report

  5  

30  Day,  Al l -­‐Cause  Readmission  Rates    

Background  –  We  assess  readmission  rates  through  two  sources.        

CMS  30-­‐Day,  All  Cause  Readmission  Rates  for  AMI  /  CHF  /  Pneumonia  –  CMS  reports  30  day,  all-­‐cause  readmission  rates  for  patients  admitted  with  AMI,  CHF,  or  pneumonia.    Because  they  have  complete  claims  data,  they  are  able  to  identify  Medicare  patients  readmitted  to  any  hospital  in  the  country.    CMS  calculates  this  data  once  annually  and  reports  it  publicly  at  www.hospitalcompare.hhs.gov.    Current  measures  are  for  July  2008  through  June  2011.        CMS  Hospital  Readmissions  Reduction  Program  for  AMI  /  CHF  /  Pneumonia  –  CMS  has  been  calculating  and  publicly  reporting  the  readmission  measures  for  Hospital  Inpatient  Quality  Reporting  since  2009  (see  above).    The  2010  Affordable  Care  Act  (ACA)  requires  the  Secretary  of  Health  and  Human  Services  to  establish  a  Hospital  Readmissions  Reduction  Program  that  would  reduce  CMS  Inpatient  Prospective  Payment  System  (IPPS)  payments  beginning  October  1,  2012.    The  ACA  further  requires  the  adoption  of  the  30-­‐day  Risk  Standardized  Readmission  measures  for  AMI,  CHF  and  Pneumonia.    To  comply  with  these  requirements  CMS  has  calculated  an  Excess  Readmission  Ratio  that  will  be  used  to  determine  payment  adjustment  for  each  eligible  hospital.    The  data  period  for  calculating  the  Excess  Readmission  Ratio  in  the  first  year  will  be  based  on  July  1,  2008  through  June  30,  2011.            Premier  30-­‐Day,  All  Cause  Readmission  Rates  –  We  assess  system,  facility,  and  DRG  business  line  level  data  for  30  day,  all-­‐cause  readmissions  to  the  same  facility  utilizing  the  Premier  Clinical  Advisor  database.    A  readmission  rate  index  is  calculated  that  represents  a  risk-­‐adjusted  measure  of  the  observed  readmission  rate  divided  by  the  expected  readmission  rate.    A  higher  than  expected  readmission  rate  can  be  an  indicator  of  poor  quality  care  in  the  hospital,  premature  discharge  from  the  hospital,  or  problems  within  the  ambulatory  care  delivery  system.          A  note  on  measurement:    The  collection  and  interpretation  of  this  data  is  complex.    Healthcare  data  is  dynamic  and  a  readmission  rate  can  be  one  of  the  most  variable  measures  in  healthcare  systems  due  to  a  variety  of  factors.    In  order  to  assess  readmission  rates,  the  medical  record  and  coding  of  the  care  provided  must  be  completed  for  both  the  first  and  second  admission.    Electronic  data  queries  will  capture  a  readmission  only  after  the  patient  has  been  discharged  a  second  time.    Thus  if  a  patient  has  a  long  stay  in  the  hospital  during  his  second  admission  it  could  potentially  be  at  least  several  months  before  the  data  query  will  capture  and  include  that  patient’s  readmission  in  the  data  results.    For  this  reason,  the  readmission  rate  for  any  given  quarter  may  increase  over  time  as  more  cases  are  identified.    Thus,  the  readmission  rate  is  continually  updated  as  ‘new’  patients  are  captured  in  the  data  reports.    Additionally,  current  methods  do  not  allow  the  capture  of  patients  readmitted  to  other  facilities.      

 GHS  Goal  –  Our  goal  is  to  have  our  readmission  index  be  statistically  better  than  expected.      

Page 6: January 2013 Quality and Safety Narrative Report

  6  

GHS  Results  CMS  30-­‐Day,  All  Cause  Readmission  Rates  for  AMI  /  CHF  /  Pneumonia    –    Current  results  on  Hospital  Compare  reported  in  June  2012  are  for  July  2008  through  June  2011.    Our  readmission  rates  for  all  3  populations  at  Greenville  Memorial  Hospital  (GMH)  continue  to  improve  slightly  compared  to  results  from  the  previous  year.    For  the  fourth  consecutive  year  GMH  has  rated  better  than  the  U.S.  national  average  in  AMI  and  CHF.    GMH  was  the  only  hospital  in  South  Carolina  to  achieve  this  ranking  for  AMI  and  only  one  of  four  South  Carolina  hospitals  to  achieve  this  “better  than”  ranking  in  CHF.    Additionally,  GMH  has  the  8th  lowest  overall  30  day  readmission  rate  for  CHF  in  the  entire  country.    For  Pneumonia  GMH  rated  no  different  from  the  U.S.  national  average.    Hillcrest  and  Greer  Memorial  Hospitals  are  statistically  no  different  from  the  national  average  for  CHF,  AMI  and  Pneumonia.    Note  that  as  the  population  becomes  smaller  around  a  specific  disease,  it  is  very  difficult  to  show  statistical  significance.        CMS  Hospital  Readmissions  Reduction  Program  for  AMI  /  CHF  /  Pneumonia  –  The  Excess  Readmission  Ratio  is  a  measure  of  relative  performance.    If  a  hospital  performs  better  than  an  average  hospital  that  admitted  similar  patients  (that  is,  patients  with  similar  risk  factors  for  readmission  such  as  age  and  co-­‐morbidities),  the  ratio  will  be  less  than  1.0000.    If  a  hospital  performs  worse  than  average,  the  ratio  will  be  greater  than  1.0000.    Results  for  Greenville  Memorial,  Hillcrest  Memorial,  and  Greer  Memorial  for  all  3  clinical  populations  were  less  than  1.0000.    GMH’s  Excess  Readmission  Ratio  for  Heart  Failure  was  0.7959  which  is  consistent  with  other  benchmarking  initiatives  that  illustrate  top  performing  status.        Premier  30-­‐Day,  All  Cause  Readmission  Rates  –  Our  system  wide  30  day,  all-­‐cause  readmission  rate  for  October  2011  through  September  2012  is  8.46%  and  our  readmission  rate  index  is  0.81  which  is  statistically  significantly  better  than  expected.    Readmission  rates  for  all  4  acute  care  hospitals  are  statistically  significantly  better  than  expected.    DRG  level  readmission  rate  indices  are  presented  for  Greenville  Hospital  System.    No  DRG  Business  Line  is  statistically  unfavorable  in  the  most  recent  quarter,  July  –  September  2012.      

   

AHRQ  Patient  Safety  Culture  Survey    Background  –  Key  to  Patient  Safety  is  the  development  of  an  organization  wide  culture  of  safety.    This  is  best  measured  using  the  AHRQ  Patient  Safety  Culture  Survey  tool  with  standardized  results  and  benchmarks.    AHRQ  publishes  their  benchmarks  typically  a  year  after  they  are  obtained.    We  survey  all  GHS  employees  and  physicians  each  August.    GHS  Goal  –  For  FY  2012,  our  GHS  organization  wide  goal  for  the  AHRQ  Patient  Safety  Culture  was  to  be  in  the  top  quartile  using  a  rolled  up  measure  of  the  entire  survey  tool.    AHRQ  reports  their  data  a  year  after  it  is  collected.    Thus,  the  AHRQ  benchmarks  we  used  to  set  our  goal  came  from  the  2011  AHRQ  Report  that  included  data  collected  in  2010  and  2011.    We  approximated  the  ~82nd  percentile  as  the  half-­‐way  point  between  the  75th  and  90th  percentiles  reported  by  AHRQ.          

Page 7: January 2013 Quality and Safety Narrative Report

  7  

(1) <  50th  percentile                                                              <  61.99%  or  lower  (2) 50th  to  74th  percentile                                          62.00%  to  66.99%  (3) 75th  to  ~82nd  percentile                                    67.00%  to  68.99%  (4) ~82nd  to  90th  percentile                                    69.00%  to  70.99%  (5) >  90th  percentile                                                              71.00%  or  higher      

 While  we  continue  to  strive  for  top  quartile  performance  on  the  rolled  up  and  individual  composite  scores,  the  GHS  Organization-­‐wide  Goal    for  FY  2013  is  to  increase  overall  survey  participation  to  60%  rather  than  achieve  a  specific  score.      GHS  Results  –  In  December  2008,  GHS  took  the  survey  for  the  first  time.    We  surveyed  only  clinical  staff  and  we  had  a  response  rate  of  55.2%  with  an  overall  score  of  59.8%.        In  August  /  September  2010,  GHS  again  took  the  survey,  but  this  time  did  it  electronically.    We  again  surveyed  only  clinical  staff  and  we  had  a  response  rate  of  35.7%  (2,138  /  5,996)  and  an  overall  score  of  62.4%.    This  was  a  statistically  significant  improvement  from  baseline  and  approximated  the  57th  percentile.    For  comparative  purposes,  the  national  mean  was  62%,  median  was  61%,  75th  percentile  was  67%,  and  maximum  was  85%.        The  FY  2011  survey  was  administered  from  August  15th  to  September  5th,  2011.    This  year,  rather  than  surveying  only  clinical  staff,  we  made  the  survey  available  electronically  through  a  link  on  GHSNet  to  all  GHS  employees  and  sent  the  survey  link  via  e-­‐mail  invitation  to  physicians.    This  is  how  AHRQ  usually  does  their  surveys  and  typically  results  in  lower  response  rates,  but  higher  scores.    Our  response  rate  decreased  to  27.2%  (2,742  /  10,097)  and  the  overall  score  increased  to  62.8%.        The  FY  2012  survey  was  administered  from  August  13th  to  September  3rd,  2012.    This  year  we  again  made  the  survey  available  electronically  through  a  link  on  GHSNet  to  all  GHS  employees  and  sent  the  survey  link  via  e-­‐mail  invitation  to  physicians.    Our  response  rate  did  increase  to  28.4%  (3100/10,934).    The  overall  score  for  GHS  increased  to  64.2%,  which  is  above  the  AHRQ  mean  of  63%  and  a  statistically  significant  increase  from  the  2008  baseline.    We  have  not  yet  reached  our  goal  of  top  quartile  performance  of    68%  (75th  percentile).        The  overall  score  is  made  up  of  12  domains.  The  safety  culture  is  particularly  positive  in  two  areas.    GHS  employees  have  the  perception  that  within  a  unit,  there  is  exceptional  teamwork  and  cooperation  (83.4%).    Secondly,  there  is  a  strong  perception  that  unit  manager  expectations  and  actions  promote  patient  safety  (79.2%).    Both  of  these  measures  are  in  the  top  quartile  nationally.        Previously,  we  had  identified  3  significant  areas  of  opportunity:  (1)  the  perception  of  a  punitive  culture;  (2)  handoffs  and  transitions;  and  (3)  teamwork  across  units.      

o In  2012,  of  the  three  previously  identified  opportunities,  handoffs  and  transitions  is  now  our  most  serious  challenge  with  a  score  of  38.9%  (mean  45%,  75th  51.0%).  This  is  followed  by  staffing  as  a  growing  concern  at  55.4%  (mean  57%,  75th  62%).    

Page 8: January 2013 Quality and Safety Narrative Report

  8  

 o Although  it  continues  to  be  a  challenge,  our  performance  related  to  Just  Culture  (non-­‐

punitive  culture)  has  improved  significantly  from  32.4%  to  45.0%  (mean  44%,  75th  49.0%).    This  was  our  lowest  performing  domain  in  FY  2010.    To  improve  this  score  we  provided  just  Culture  training  in  the  May  2011  Leadership  Development  retreat  and  a  break-­‐out  training  update  on  “Just  Culture”  at  our  July  2012  Leadership  Development  Retreat.      

 Detailed  reports  for  each  hospital  are  provided  in  a  separate  annual  report.      

National  Patient  Safety  Goals    

Background  –  The  Joint  Commission  (TJC)  has  established  a  number  of  National  Patient  Safety  Goals  (NPSG),  which  are  process  steps  that  should  be  implemented  to  ensure  optimal  patient  safety.    NPSGs  are  not  publicly  reported  and  they  represent  a  self-­‐audit.    Consequently,  there  is  no  national  comparative  data.    In  2011  the  GHS  audit  process  for  NPSGs  changed  when  a  new  methodology  for  data  collection  was  developed.    Previously,  compliance  was  evaluated  by  a  unit  self-­‐audit.    Data  is  now  collected  by  the  Quality  Management  Data  Collector  Nurses.        The  currently  reported  NPSGs  include  the  following:    

• NPSG  1  –  Patient  Identification:    Use  at  least  two  patient  identifiers  when  administering  medications,  blood,  or  blood  components;  when  collecting  blood  samples  and  other  specimens  for  clinical  testing;  and  when  providing  treatments  or  procedures.  The  patient's  room  number  or  physical  location  is  not  used  as  an  identifier.    Label  containers  used  for  blood  and  other  specimens  in  the  presence  of  the  patient.      Audit  methodology:  Staff  are  directly  observed  by  Quality  Management  RNs  while  performing  procedures  for  compliance  with  the  requirements  for  patient  identification  including  the  use  of  barcoding  technology.    

 • NPSG  2  –  Reporting  of  Critical  Result:    Notification  of  appropriate  providers  of  a  critical  test  

result  within  one  hour  of  the  test’s  availability.    Audit  methodology  selected:    A  list  of  critical  results  is  obtained  from  the  laboratory;  then,  a  chart  audit  is  done  for  the  documentation  and  timeliness  (one  hour  or  less  turn-­‐around  time)  of  reporting  critical  results.    

• NPSG  3.06  –  Medication  Reconciliation:    Obtain  information  on  the  medications  the  patient  is  currently  taking  when  he  or  she  is  admitted  to  the  hospital  or  is  seen  in  an  outpatient  setting  and  compare  the  medication  information  the  patient  brought  to  the  hospital  with  the  

Page 9: January 2013 Quality and Safety Narrative Report

  9  

medications  ordered  for  the  patient  by  the  hospital  in  order  to  identify  and  resolve  discrepancies.    Provide  the  patient  (or  family  as  needed)  with  written  information  on  the  medications  the  patient  should  be  taking  when  he  or  she  is  discharged  from  the  hospital  or  at  the  end  of  an  outpatient  encounter  and  explain  the  importance  of  managing  medication  information  to  the  patient  when  he  or  she  is  discharged  from  the  hospital  or  at  the  end  of  an  outpatient  encounter.    Audit  methodology:    Medical  charts  are  audited  for  evidence  of  a  completed  medication  list  on  admission;  reconciliation  of  the  medication  list;  and  medications  to  be  listed  along  with  completed  education  of  the  patient  and  family  at  discharge.    

 • NPSG  15  –  Suicide  Risk  Assessment  and  Safety:    Conduct  a  risk  assessment  that  identifies  specific  

patient  characteristics  and  environmental  features  that  may  increase  or  decrease  the  risk  for  suicide,  and  address  the  patient’s  immediate  safety  needs  and  most  appropriate  setting  for  treatment.        Audit  methodology:    On  the  day  this  NPSG  is  audited,  a  list  of  behavioral  patients  present  in  the  emergency  department  is  obtained.    A  review  of  those  patients’  chart  is  conducted  to  determine  if  the  initial  suicide  risk  assessment  was  completed,  as  well  as  evidence  of  on-­‐going  risk  assessment.      

• Universal  Protocol  /  Bedside  Time-­‐out:    Implement  a  preprocedure  process  to  verify  the  correct  procedure,  for  the  correct  patient,  at  the  correct  site,  mark  the  procedure  site  (if  applicable),  and  perform  a  time-­‐out  before  the  procedure.    Audit  methodology:    A  chart  audit  is  done  for  the  presence  of  the  completed  bedside  time-­‐out  form  on  the  charts  of  patients  who  have  had  a  bedside  procedure.      

GHS  Goal  –  Specific  organization  wide  goals  for  the  NPSGs  have  not  been  set,  but  best  practice  encourages  that  they  should  be  carried  out  100%  of  the  time.    TJC  typically  expects  90%  compliance.    During  a  recent  Joint  Commission  visit  GHSUMC  was  commended  on  performance  improvement  efforts  for  Patient  Identification  related  to  labeling  of  laboratory  specimens  at  the  bedside  as  evidenced  by  the  National  Patient  Safety  Goal  data  and  observation  during  the  survey.      GHS  Results  –  Current  quarter  results  range  between  84.1%  and  99.3%  with  an  overall  score  of  91.4%.    This  represents  a  significant  improvement  in  individual  and  overall  NPSGs  and  is  the  first  time  since  the  change  in  the  auditing  methodology  that  the  overall  score  has  exceeded  90%.      Significant  improvement  has  been  sustained  for  suicide  risk  assessment  and  critical  results  after  focused  improvement  initiatives  were  completed.    There  was  also  significant  improvement  in  medication  reconciliation  due  to  the  addition  of  educational  information  to  the  Patient  Admission  Packet  related  to  

Page 10: January 2013 Quality and Safety Narrative Report

  10  

the  maintenance  of  the  discharge  medication  list  in  order  to  comply  with  the  requirement.  This  was  completed  at  the  end  of  the  second  quarter.        

 Anticoagulat ion  Therapy  and  Monitoring  

 Background – Bleeding  from  anticoagulant  use  has  been  recognized  as  a  prevalent  complication  and  an  important  cause  of  morbidity  and  mortality.    For  this  reason  Joint  Commission  added  National  Patient  Safety  Goal  03.05  that  required  the  development  and  use  of  hospital  policies  to  help  reduce  the  likelihood  of  harm  to  patients  receiving  therapeutic  anticoagulation  or  long-­‐term  prophylaxis.    GHS  initiatives  began  January  1,  2009.    In  2010,  a  multidisciplinary  team  from  Greenville  Memorial  collaborated  with  industrial  engineers  from  Clemson  University  to  analyze  and  direct  further  efforts  to  reduce  adverse  drug  events  associated  with  warfarin  (Coumadin).    Numerous  changes  to  processes  were  made.    The  International  Normalized  Ratio  (INR)  is  used  to  monitor  the  extent  of  anticoagulation.    It  may  take  several  days  of  dosing  for  the  INR  to  reach  therapeutic  values.    As  part  of  this  initiative,  specific  metrics  were  developed:  

• Measure  1:    The  percent  of  patients  with  an  INR  between  2.0  to  3.5  representing  optimal  levels  of  anticoagulation.  

• Measure  2:    The  percent  of  patients  with  an  INR  in  the  “Critical  Value  Range”,  which  is  defined  as  being  greater  than  or  equal  to  5.0  and  carries  with  it  a  significantly  increased  risk  of  bleeding.      

• The  denominator  is  the  count  of  INR  values  for  hospitalized  patients  who  received  Warfarin.        Since  2010,  GHS  pharmacists  have  been  increasingly  involved  with  education  and  monitoring  of  patients  on  warfarin  in  multiple  outpatient  clinics  to  reduce  admissions  with  INR  values  >  5.  

 Goal  –  Reduce  the  likelihood  of  patient  harm  related  to  major  bleeding  events  as  evidenced  by  an  increase  in  the  percent  of  INRs  within  the  therapeutic  range  and  a  decrease  in  the  percent  of  INRs  at  or  over  the  critical  range.        Results  –  The  data  points  reflect  a  rolling  12-­‐month  period,  trending  January  2009  forward.    The  percentage  of  therapeutic  INR  values  for  January-­‐December  2012  is  35.80%,  which  continues  to  trend  positively.    The  percentage  of  critical  INR  values  is  1.91%,  showing  a  less  favorable  trend  from  the  preceding  year  and  being  equivalent  to  Jan-­‐Dec  2009.    It  is  notable  that  the  Greenville  Memorial  acute  care  population  being  treated  with  warfarin  has  diminished  by  332  as  compared  to  calendar  year  2011,  and  626  fewer  than  CY  2010.    The  denominator  for  total  INR  values  is  2,813  less  than  CY  2011.    Some  of  the  decline  in  utilization  can  be  related  to  newer  oral  anticoagulants  approved  by  the  FDA  in  2010  and  added  to  the  GMH  formulary  in  May  2011.          

Page 11: January 2013 Quality and Safety Narrative Report

  11  

Hospital  Acquired  Condit ions  (HACs)    

Background  –  The  Center  for  Medicare  and  Medicaid  Services  (CMS)  recently  adopted  eight  of  the  ten  Hospital  Acquired  Condition  measures  as  part  of  their  Pay-­‐for-­‐Reporting  requirements.    This  initial  set  of  eight  measures  will  be  publicly  reported  on  the  CMS  Hospital  Compare  site  by  June  2011  as  a  downloadable  file.    The  selected  measures  were  established  in  collaboration  with  the  CDC  and  other  external  agencies  to  determine  conditions  or  events  which  were  considered  serious  and  reasonably  preventable  through  application  of  evidence-­‐based  guidelines.    The  conditions  are  identifiable  through  claims  data  for  Medicare  fee-­‐for-­‐service  patients  only.    Identification  of  inpatients  with  a  HAC  is  determined  through  the  use  of  qualifying  ICD  diagnostic  codes  and  qualifying  Present  on  Admission  (POA)  codes.    In  addition  CMS  has  proposed  to  include  the  eight  HAC  measures  as  part  of  its  Value  Based  Purchasing  Initiative  in  2014,  but  recently  suspended  this  initiative  for  one  year.          The  eight  Hospital  Acquired  Conditions  that  CMS  will  begin  to  publicly  report  are:  

1. Retained  Foreign  Object  after  surgery  2. Air  Embolism  3. Blood  Incompatibility  4. Pressure  Ulcer  5. Falls  and  Trauma  6. Vascular  Catheter-­‐Associated  Infection  7. Catheter-­‐Associated  Urinary  Tract  Infection  8. Poor  Glycemic  (blood  sugar)  Control  

 There  is  significant  concern  regarding  the  accuracy  of  these  measures.    They  are  all  developed  exclusively  from  claims  data  which  is  subject  to  errors  in  documentation  and  coding.    In  many  situations,  the  HAC  data  is  not  correlating  with  much  more  specific  data  at  GHS  that  is  obtained  using  detailed  condition  definitions  and  chart  audits.    Additionally,  in  some  circumstances,  there  are  medically  justified  reasons  for  a  HAC  to  occur.    They  may  not  be  100%  preventable.    GHS  Goal  –  No  goal  has  been  set  at  this  time  for  HACs.    Ultimately  our  goal  will  be  to  minimize  the  number  of  HAC’s  for  all  eight  measures  across  the  system.          GHS  Results  –  CMS’s  initial  HAC  report  covers  the  time  frame  July  2009  –  June  2011.    All  HAC  measures  for  Greenville  Memorial  Hospital  were  lower  than  the  national  rates  except  for  Foreign  Object  Retained  after  surgery  (1  occurrence),  Catheter-­‐Associated  Urinary  Tract  Infection  (GMH  rate  is  0.910  compared  to  a  national  rate  of  0.358)  and  Poor  Glycemic  Control  (two  events  resulted  in  a  GMH  rate  of  0.087  compared  to  a  national  rate  of  0.058).    Greer  Memorial  Hospital  had  one  event  for  the  measure  “Falls  and  Trauma”  resulting  in  a  rate  of  0.36  compared  to  a  national  benchmark  rate  of  0.527.    Hillcrest  Memorial  Hospital  also  had  one  event  for  the  measure  “Falls  and  Trauma”  resulting  in  a  rate  of  0.608  compared  to  a  national  benchmark  rate  of  0.527.    All  other  HAC  measures  for  Greer  and  Hillcrest  showed  zero  events.    Patewood  Memorial  Hospital  had  no  identified  HAC’s  during  this  time  frame.        

Page 12: January 2013 Quality and Safety Narrative Report

  12  

AHRQ  Patient  Safety   Indicators  (PSIs)    

Background  –  A  method  of  assessing  inpatient  patient  safety  and  complication  events  is  to  use  the  Agency  for  Healthcare  Research  and  Quality  (AHRQ)  Patient  Safety  Indicators  (PSIs).    The  PSIs  are  a  set  of  measures  that  provide  perspective  on  hospital  quality  of  care  using  hospital  administrative  (claims)  data.    The  indicators  are  used  to  screen  for  potential  adverse  events  occurring  during  hospitalization  following  surgeries,  procedures  and  childbirth.    They  are  based  on  evidence  based  medicine  and  use  complex  algorithms  that  are  risk  adjusted.    While  the  PSIs  were  intended  for  internal  screening  to  identify  potential  areas  of  improvement  opportunity,  they  are  now  frequently  being  used  to  rate  the  quality  and  safety  of  care  delivered  by  hospitals.    At  this  time  CMS  is  publicly  reporting  on  their  Hospital  Compare  website  the  following  seven  AHRQ  PSI  indicators:      Patient  Safety  Indicator        

• Iatrogenic  Pneumothorax      • Post  op  PE  or  DVT          • Post  op  Wound  Dehiscence    • Accidental  Puncture  or  Laceration  • Death  among  Surgical  Inpatients  with  Serious  Treatable  Conditions  • Post  op  Respiratory  Failure  • Complications/Patient  Safety  for  Selected  Indicators  (Composite  Score)  

 There  is  significant  concern  regarding  the  accuracy  of  these  measures.    They  are  all  developed  exclusively  from  claims  data  which  is  subject  to  errors  in  documentation  and  coding.        GHS  Goal  –  No  goal  has  been  set  as  these  are  screening  tools.    For  each  PSI,  we  would  like  the  actual  measure  to  be  lower  than  the  comparative  benchmark.          GHS  Results  –  UHC  and  Premier  provide  external  benchmarks  for  the  PSI’s  and  thus  we  are  able  to  track  our  performance  throughout  the  year.    Current  data  through  September  2012  continues  to  trend  favorably  with    fewer  opportunities  for  Greenville  Memorial.    Post-­‐operative  pulmonary  embolus  or  deep  vein  thrombosis  and  birth  injury  to  neonates  both  are  statistically  better  than  expected  compared  to  a  benchmark.    Opportunities  include  post-­‐operative  respiratory  failure,  post-­‐operative  sepsis  and  obstetrical  trauma  during  vaginal  delivery  without  instruments.    No  outliers  exist  at  Greer,  Hillcrest  or    Patewood  relative  to  Premier  external  benchmarks.          A  Lean  Six  Sigma  project  was  launched  in  January  2012  to  develop  a  methodology  to  improve  AHRQ  Patient  Safety  Indicators  (PSI).  The  initial  focus  was  around  the  PSI  ‘Accidental  Puncture  or  Laceration;  however,  the  project  also  focused  on  PSI  for  ‘Post  Op  Hemorrhage  &  Hematoma’  and  ‘Birth  Trauma  Injury  to  Neonate’.      A  significant  finding  of  this  work  is  that  the  Accidental  Puncture  or  Laceration  PSI  

Page 13: January 2013 Quality and Safety Narrative Report

  13  

was  driven  primarily  by  documentation  and  coding  issues  rather  than  clinical  care.    After  addressing  both  clinical  and  documentation  issues  Greenville  Memorial  has  achieved  top  quartile  rates  compared  to  UHC  benchmarks  over  the  past  four  quarters.        An  AHRQ  PSI  that  we  have  begun  to  trend  internally  even  though  it  is  not  yet  publicly  reported  is  the  rate  of  pressure  ulcers.    Reviewing  our  results  compared  to  the  UHC  benchmark  shows  that  the  overall  pressure  ulcer  rate  at  Greenville  Memorial  is  at  or  below  benchmark.      The  pressure  ulcer  process  improvement  core  team  is  developing  a  project  charter  and  preparing  for  the  inaugural  team  meeting  in  February.        

IHI  Global  Tr igger  Tool    

Background  –  Traditional  efforts  to  detect  adverse  events  have  focused  on  voluntary  reporting  of  events  by  healthcare  providers  and  tracking  and  trending  of  errors.    Public  health  researchers  have  established  that  only  10  to  20  percent  of  errors  are  ever  reported  and,  of  those,  90  to  95  percent  cause  no  harm  to  patients.  The  IHI  Global  Trigger  Tool  for  Measuring  Adverse  Events  was  implemented  in  2003,  providing  a  method  for  accurately  identifying  events  (harm)  and  measuring  the  rate  of  adverse  events  over  time.  “Harm”  is  defined  as  “unintended  physical  injury  resulting  from  or  contributed  to  by  medical  care  that  requires  additional  monitoring,  treatment  or  hospitalization,  or  that  results  in  death”.  The  concept  is  to  move  from  a  focus  on  error  and  whether  it  was  preventable  or  not,  to  the  measurement  of  global  institutional  harm,  whether  preventable  or  not.    The  process  involves  a  retrospective  review  of  a  random  sample  of  inpatient  medical  records  using  “triggers”  or  clues,  to  identify  possible  adverse  events.        GHS  Goal  –  A  goal  will  be  set  after  a  establishing  a  baseline  since  IHI  recommends  obtaining  at  least  twelve  data  points  prior  to  performing  data  analyses.          GHS  Results  –  The  Quality  Management  Department  began  using  the  IHI  Global  Trigger  tool  in  January  2012,  reviewing  20  random  charts  per  month.      Based  on  four  quarters  of  data  67  patients  had  82  harm  events  that  were  identified  through  chart  reviews  and  were  corroborated  by  the  appropriate  Vice  Chairs  of  Quality.    One  way  to  present  the  data  is  in  a  run  chart  that  displays  the  number  of  ‘Adverse  Harm  Events  per  1000  patient  days’.    GHS  results,  which  includes  Greenville,  Hillcrest,  Greer  and  Patewood  Memorial  Hospitals,  range  from  50.2  to  79.9  harm  events  per  thousand  patient  days  per  quarter.    Florida  Hospital  System’s  historical  average  was  selected  as  a  benchmark  since  it  is  also  a  multi-­‐facility  system  similar  to  GHS  and  has  years  of  experience  with  the  IHI  Global  Trigger  Tool.        

 Event  Reporting  

 Background  –  Critical  to  the  ability  to  improve  quality  and  prevent  adverse  events  is  the  need  to  identify  errors  and  near  misses,  analyze  and  understand  opportunities  for  improvement  and  implement  

Page 14: January 2013 Quality and Safety Narrative Report

  14  

targeted  improvement  initiatives.    This  identification  and  measurement  of  actual  and  potential  adverse  events  is  critical  to  the  development  of  a  safety  culture  and  a  high  reliability  organization.    The  search  for  opportunities  for  improvement  comes  from  data  across  a  spectrum  that  includes  patient  complaints,  reported  unsafe  conditions,  near  misses  and  adverse  events,  the  investigation  of  adverse  events  and  malpractice  litigation.        Unsafe  conditions  represent  issues  that  present  the  potential  for  patient  safety  issues  if  not  corrected  and  include  such  things  as  computer  system,  equipment  safety,  infrastructure  failure  and  security  issues.    They  are  not  patient  specific.    The  ability  to  proactively  identify  and  trend  such  issues  via  the  event  reporting  system  has  been  in  place  since  2011.      Near  misses  and  adverse  events  both  relate  to  the  care  of  a  specific  patient.    Near  misses  are  potential  events  that  were  caught  and  prevented  prior  to  the  patient  being  involved.    An  adverse  event  occurs  when  the  event  or  care  did  involve  the  patient.    The  adverse  event  may  or  may  not  have  caused  any  patient  harm.    It  is  reported  in  the  literature  that  typically  only  5  to  10%  of  errors  are  actually  reported  in  hospitals.    This  is  also  consistent  with  baseline  data  from  other  high  risk  industries.    Thus,  there  is  a  significant  need  to  increase  the  rate  of  reporting  of  all  errors  in  order  to  detect  and  mitigate  potential  harm  to  patients.    GHS  converted  to  University  HealthSystem  Consortium’s  (UHC)  Patient  Safety  Net  (PSN)  for  event  reporting  in  late  December,  2010.    This  web-­‐based  tool  provides  a  mechanism  to  identify,  catalogue  and  analyze  events  and  unsafe  conditions,  which  can  then  be  systematically  corrected  to  improve  outcomes  and  prevent  patient  injury.                GHS  Goal  –  We  measure  the  reporting  of  errors  and  adverse  events  as  a  rate  for  inpatient  settings  (number  of  events  reported  per  1000  patient  days)  and  as  a  rate  for  outpatient  settings  (number  of  events  reported  per  10,000  procedures).    The  current  goal  is  set  at  the  75th  percentile  of  Event  Reporting  compared  to  comparable  size  hospitals  in  the  UHC  database.      Thus,  the  goal  is  for  the  GHS  overall  Inpatient  Event  Reporting  Rate  to  be  at  or  above  40.18  reports  per  1000  patient  days  for  each  of  our  facilities.    No  benchmark  has  been  established  for  outpatient  event  rate  as  published  comparison  data  is  not  available.      GHS  Results  –  Current  results  are  for  the  forth  quarter  of  CY  2012.      

Frequency  –  There  has  been  a  slight  decrease  in  patient  event  rate  during  the  current  quarter.  Event  reporting  was  not  included  in  organizational  goals  for  FY2013,  which  might  contribute  to  inconstancy  of  reporting  events  beginning  October,  2012.  As  a  system,  GHS  continues  to  be  below  the  UHC  75th  percentile  of  40.18  per  1000  patient  days.  The  baseline  event  reporting  rate  in  2010  was  11.9.    The  rate  for  December,  2012  was  26.6,  a  slight  improvement  over  prior  month.  The  volume  of  event  reporting  hit  a  high  of  32.9  in  1QTR  12  but  has  declined  11%  since  that  time  to  a  current  rate  in  4QTR12  of  29.25.      

Page 15: January 2013 Quality and Safety Narrative Report

  15  

Severity  –  For  the  quarter  reported,  the  rate  of  events  with  moderate  to  severe  injury  remained  a  small  percent  of  the  total  reports  and  is  in  line  with  prior  months.    Inpatient  events  with  harm  in  4QTR12  remained  at  2.3  per  1000  patient  days,  basically  unchanged  from  prior  quarter.        Type  –  Event  type  allows  reporting  of  patient  and  visitor  events,  as  well  as  unsafe  conditions.    The  most  common  event  types  reported  include:  

1) Laboratory  test  (25%  -­‐  essentially  unchanged  from  the  prior  quarter)  2) Falls  dipped  slightly  from  the  prior  quarter,  constituting  13%  of  total  events.    It  appears  

that  fall  frequency  remains  fairly  steady.    The  GHS  fall  rate  of  4.0  for  4QTR  2012  shows  a  downward  trend  for  the  year.    

3) Medication  related  events    were  down  8%  from  the  prior  quarter,  constituting  13%  of  all  event  types.    

4) Complications  of  Care  (unanticipated,  nonsurgical)  increased  14%  this  quarter.      6)      Staff  have  been  working  arduously  to  educate  front  line  reporters  to  select  the  

appropriate  event  category  when  reporting.  This  has  reduced  the  percentage  of  “other”  event  report  type  selection  which  fell  13%  this  quarter.    

 Of  critical  note  is  that  the  relative  proportions  of  types  of  events  may  not  be  reliable.    There  is  a  significant  bias  on  the  part  of  staff  relative  to  past  training  to  report  some  types  of  events  and  not  others.    For  example,  staff  are  well  trained  that  patient  falls  always  need  to  be  reported.    This  is  in  contrast  to  other  types  of  events  which  staff  may  not  be  as  inclined  to  report.  

 Falls  –  The  Fall  rate  has  remained  fairly  consistent  over  the  past  five  quarters.  The  goal  for  falls  has  been  set  at  3.18  /  1000  patient  days  and  is  based  on  the  UHC  benchmark  data  for  falls.  The  fall  rate  for  North  Greenville  (2.2)  has  been  below  the  goal  for  three  consecutive  quarters.  Hillcrest  (2.4)  fell  below  that  goal  in  the  most  recent  two    quarters.    Greer  dipped  below  the  goal  in  4QTR12  with  2.9.  All  other  facilities  are  above  the  goal  rate  of  3.18.  The  fall  rate  for  the  current  quarter  for  Greenville  Memorial  is  3.3.  Marshall  Pickens  and  The  Cottages  at  Brushy  Creek  increased  this  quarter  to  4.9.      Patewood  (9.7)  and  Roger  C.  Peace  (7.5)  have  the  highest  fall  rates.      

 In  an  effort  to  improve  performance,  the  Falls  Prevention  Program  underwent  a  significant  change  in  2011-­‐2012.  Initiatives  over  the  past  12  months  include:    

•  Complete  revision  of  Fall  Prevention  Policy;    • A  new  Fall  Risk  Assessment  and  the  Morse  Fall  Scale  was  implemented  with  

interventions  based  on  scoring;        • Implementation  of  post  Fall  Huddles  to  discuss  the  causes  of  the  fall  and  interventions  

needed  to  prevent  another  fall.      

Page 16: January 2013 Quality and Safety Narrative Report

  16  

The  Falls  Committee  continues  to  work  with  Marketing  on  a  system-­‐wide  marking  initiative  for  Falls  prevention.  That  team  is  working  with  various  units  /  facilities  to  provide  more  intensive  interventions  and  staff  engagement,  as  well.          

   

A lar is  Guardrai ls®  Suite    Alaris  Guardrails®  Suite      Background  -­‐  Alaris  infusion  pumps  for  medications  are  equipped  with  Guardrails®  Suite  which  notifies  practitioners  when  an  infusion  amount  is  programmed  outside  clinical  best  practices  for  drug  delivery.    One  quality  measure  is  usage  of  the  Guardrails®  technology  or  the  percentage  of  infusions  that  take  advantage  of  hospital  defined  clinical  best  practices  for  drug  delivery.      Goal  -­‐  GHS  has  elected  a  goal  of  usage  of  Guardrails  technology  in  70%  of  infusions.          Results  -­‐  During  the  current  period  (October,  2012)  Guardrails  usage  was  74%,  which  is  a  decrease  from  75%  in  the  previous  period.  To  improve  usage,  Pharmacy  continues  to  evaluate  profiles  for  medications  or  devices  not  regularly  using  the  technology.  Nursing  also  continues  education  on  the  equipment.      Alaris  Guardrails  Suite  Overrides      Background  -­‐  Another  measure  is  the  percent  of  times  staff  override  the  alerts  of  the  Guardrails®  Suite.  When  a  pump  is  programmed  for  a  specific  medication  to  infuse  at  a  certain  rate,  and  the  rate  falls  outside  clinical  best  practices,  the  pump  signals  an  audible  alarm.  The  nurse  can  elect  to  cancel  the  infusion,  reprogram  the  infusion,  or  override  the  alert  and  start  the  pump.      Goal  –  GHS  has  not  specified  a  goal  for  Guardrails  Suite®  overrides.    Results  -­‐  During  October,  2012,  overrides  increased  68%,  with  129  overrides  per  1,000  Guardrails  Infusion  Starts  compared  to  61  in  the  prior  month.        There  has  been  a  steady  increase  in  trendline  for  “Good  Catches”  from  Guardrails  Suite®  usage    over  the  past  6  months  .  The  number  of  severe  harm  averted  through  use  of  Alaris  Good  Catches  was  25  in  the  month  of  October,  2012.          

Infect ion  Prevention    

Background  –  GHS  has  a  comprehensive  Infection  Prevention  and  Control  Program  which  encompasses  prevention  and  control  practices,  targeted  ongoing  infection  surveillance,  and  process  

Page 17: January 2013 Quality and Safety Narrative Report

  17  

improvement  to  minimize  infection  risk.    Targeted  healthcare  associated  infections  are  also  publicly  reported  in  South  Carolina  and  are  displayed  on  the  SC  DHEC  web-­‐site.    For  FY  2013,  top  priorities  include,  Catheter  associated  urinary  tract  infection  (  CAUTI),    hand  hygiene,  central  line  associated  bloodstream  infection  (CLABSI),  ventilator-­‐associated  pneumonia  (VAP),  surgical  site  infections  (SSI)    and  multi-­‐drug  resistant  organisms  (MDRO).    CAUTI  surveillance    was  expanded  during  2012  and  consistent  reporting  on  this  device  related  infection  will  be  a  component  of  board  reports  during  this  year.      Physician  led,  collaborative  teams  are  established  to  facilitate  infection  risk  reduction  for  each  of  the  priority  areas.    This  report  does  not  reflect  all  of  the  surveillance  and  work  of  the  Infection  prevention  program,  but  focuses  on  the  top  priorities.        GHS  Goal  –  Strive  to  eliminate  infections.    Infection  rate  targets  are  established  annually  to  promote  continuous  improvement.    The  benchmark  is  obtained  from  the  National  Healthcare  Safety  Network  (NHSN),  a  national  surveillance  program  sponsored  by  the  CDC,  in  which  GHS  participates.    There  are  no  national  benchmarks  for  hand  hygiene  and  multi-­‐drug  resistant  organisms.    Targets  were  established  for  these  infections  based  on  internal  data.  

   

Hand  Hygiene    Background  –  Hand  Hygiene  remains  the  hallmark  of  infection  prevention  and  has  been  a  GHS  organization  wide  goal  for  the  past  three  years  (  2010–2012).  Compliance  rates  around  the  country  typically  run  around  30%  to  70%.        Most  hospitals  in  the  United  States  teach  the  hand  hygiene  method  of  “In  and  Out”.    This  involves  the  cleaning  of  hands  prior  to  (entry  to  the  room)  and  after  (exit  from  the  room)  patient  care.    There  are  significant  concerns  that  this  strategy,  while  helpful,  is  not  optimal.    Significant  opportunities  to  re-­‐contaminate  cleansed  hands  exist  once  the  provider  has  entered  the  patient’s  room.    Consequently,  GHS  made  a  decision  to  use  the  much  more  stringent  World  Health  Organization’s  “Five  Moments  of  Hand  Hygiene”  methodology  that  requires  the  provider  to  clean  their  hands  before  and  after  patient  contact  as  well  as  before  and  after  contact  with  objects  within  the  patient’s  room.    We  believe  the  5  moments  are  more  scientifically  based  and  important  as  we  have  documented  the  known  transmission  of  infection  to  patients  from  bacteria  present  in  their  environment  in  their  room.    Washing  hands  only  on  entry  and  exit  from  the  room  will  not  prevent  these  episodes  of  infection.    There  are  no  national  benchmarks  for  hand  hygiene  compliance,  but  the  literature  suggests  a  critical  target  of  90%.      Over  the  past  three  years,  GHS  has  employed  an  organization-­‐wide  campaign  “Germ  Warfare,  Join  the  Battle”.    Using  this  campaign,  GHS  was  able  to  reach  and  sustain  the    target  of  90%  hand  hygiene  compliance  measured  by  direct  observation  18  months  ahead  of  schedule.        Two  dedicated  nurses  have  provided  direct  observations  for  measurement  with  close  attention  made  to  their  inter-­‐rater  reliability  and  validity.    Numerous  strategies  have  been  used  including  the  addition  of  more  hand  hygiene  agent  dispensers  to  promote  hand  hygiene  by  staff  and  visitors  and  training  of  staff  

Page 18: January 2013 Quality and Safety Narrative Report

  18  

based  on  surveillance  and  focus  group  findings.    At  the  same  time,  GHS  has  been  collaborating  with  DebMed,  a  hand  hygiene  company  to  develop  an  electronic  method  to  monitor  hand  hygiene.      The  electronic  methodology  affords  a  means  of  24/7  monitoring  and  with  the  appropriate    denominator  (average  number  of  anticipated  hand  hygiene  opportunities)  can  establish  a  good  surrogate  for  a  hand  hygiene  compliance  rate).    We  are  in  the  process  of  validating  this  electronic  methodology  and  transitioning  the  organization  to  its  use.      Comments  on    observation  methods.  The  classic  method    for  hand  hygiene  monitoring  is  direct  observation    using    “secret  shoppers”  unknown  to  the  healthcare  workers.    Because  these  observers  cannot  necessarily  observe  care  in  the  patient’s  room,  they  usually  are  limited  to  measuring  hand  hygiene  only  when  the  healthcare  worker  enters  and  leaves  the  room.    This  is  the  methodology  used  by  Novant  when  it  was  able  to  achieve  a  90%  compliance  rate  over  3  years.    It  also  is  the  methodology  we  used  to  identify  the  baseline  of  53.8%  compliance  in  June  to  September  2009.    A  second  method  of  direct  observation    is  to  have  the  observer  introduce  themselves  to  the  healthcare  worker  and  follow  them  into  the  room.    We  are  currently  using  this    direct  observation  method  because  the  WHO  “Five  Moments  of  Hand  Hygiene”  method  does  not  easily  allow  for  direct  observations  by  an  unknown  observer  within  patient  care  rooms.    The  down  side  to  this  method  is  its  complexity  and  the  introduction  of  the  Hawthorne  Effect,  i.e.  compliance  increases  when  the  healthcare  worker  knows  they  are  being  observed.    Thus,  the  two  methods  are  both  valid,  but  likely  will  deliver  different  compliance  rates.    A  critical  factor  is  to  measure  consistently.    At  GHS,  we  are  engaged  in  a  significant  research  study  around  hand  hygiene  compliance.    The  research  being  performed  here  centers  around  an  electronic  method  to  identify  the  number  of  times  a  healthcare  worker  uses  hand  gel  or  soap  during  a  patient  encounter.    We  have  developed  statistical  models  to  identify  the  average  number  of  opportunities  a  healthcare  worker  should  clean  their  hands  based  on  the  WHO  5  moments  of  hand  hygiene  during  a  patient  encounter.    Thus,  the  combination  of  use  of  hand  cleansing  agent  (numerator)  divided  by  the  expected  opportunities  for  hand  cleansing  (denominator)  provides  us  with  an  index  to  measure  hand  hygiene  in  real  time  and  across  many  different  units  every  shift.    We  are  in  the  analysis  phase  of  a  validation  study  for  the  electronic  monitoring  system.        The  key  take  away  is  that  none  of  the  methods  is  capable  of  determining  the  actual  compliance  rate  across  the  organization.    Thus,  the  absolute  compliance  rate  is  not  as  important  as  the  trend  towards  improvement.      GHS  Goal(s)  –    1.  Quarterly  direct  observation  compliance  rate(s)  of  90%  or  greater.    2.  Transition  to  electronic  monitoring  using  validated  methodology.    3.  Initiate  a  unit  based  approach  to  identifying  barriers  to  hand  hygiene  and  take    steps  to  improve  compliance.    GHS  Results  –  Hand  hygiene  compliance  rates  by  direct  observation  continue  to  remain  above  90%.    For  the  October  to  December  2012  quarter,  the  compliance  rate  is  93.7%.    The  data  indicates  that  

Page 19: January 2013 Quality and Safety Narrative Report

  19  

healthcare  providers  clean  their  hands  for  4  of  the  5  moments  between  92.9%  and  96.3%  of  the  time.    The  opportunity  for  improvement  is  before  aseptic  /  clean  procedures  at  82.5%  and  probably  due  to  the  misconception  that  hand  hygiene  is  not  required  when  gloves  are  worn.    There  continues  to  be  gradual  improvement  in  the  areas  of  opportunity.    Nursing  staff  (nurses  and  technicians)  and  therapy  staff  are  more  likely  to  clean  hands  than  other  healthcare  provider  groups.        During  2012,  we  initiated  the    electronic  form  of    hand  hygiene  monitoring  on  targeted  units  at  GMH    and  on  the  North  Greenville  and  Hillcrest  campuses;    unit-­‐specific  compliance  indices  are  being  assessed  in  beta-­‐testing  and  unit  managers  are  being  educated  on  their  meaning.    As  expected,  the  hand  hygiene  compliance  indices  via  electronic  monitoring  are  lower  than  compliance  rates  via  direct  observation  given  that  staff’s  behavior    changes  when  being  directly  observed  (e.g.  Hawthorne  Effect).    As  part  of  the  hand  hygiene  validation  study  we  have  confirmed  that  actual  hand  hygiene  compliance  is  roughly  20-­‐30%  lower  than  “direct-­‐observation”  compliance  and  that  the  electronic  monitoring  system  provides  the  best  estimate  of  actual  compliance.    The  organization  will  continue  to  transition  the  majority  of  units  to  electronic  monitoring  over  the  next  year  and  we  will  be  publishing/  presenting  our  research  nationally.    

 Surgical  S ite   Infections  (SSIs)  

 Background  –  We  track  a  number  of  surgical  site  infection  rates  which  are  required  by  South  Carolina  law  to  be  publicly  reported  on  the  DHEC  website.    The  data  in  this  report  is  presented  in  terms  of  the  Standardized  Infection  Ratio  (SIR),  which  is  a  statistical  ratio  of  the  observed  infection  rate  divided  by  the  expected  infection  rate.    The  confidence  intervals  of  each  SIR  must  cross  1.0.    SIRs  above  1.0  demonstrate  a  worse  than  targeted  infection  rate,  while  those  below  1.0  are  better  than  targeted.        NHSN  has  recently  changed  the  methodology  for  risk  adjustment  of  SSIs  to  include  all  procedure-­‐level  data  collected  on  each  patient  (i.e.,  patient  age,  gender,  duration  of  surgery,  diabetes,  trauma,  etc.).    The  prior  risk-­‐adjustment  method  was  based  solely  on  the  ASA  (American  Society  of  Anesthesiologists)  physical  status  classification  system  (i.e.,  1=normal  healthy  patient,  …,  4=severely  ill  patient).    This  new  methodology  represents  a  significant  improvement  in  risk-­‐adjustment.    NHSN  used  the  data  from  2006-­‐2008  to  derive  the  new  risk  adjustment  models  and  then  applied  them  to  data  from  2009  forward.    Four  surgical  procedures  are  presented  with  the  new  method  for  determining  expected  numbers  of  infections  -­‐-­‐  coronary  bypass,  abdominal  hysterectomy,  hip  replacement  and  knee  replacement.    All  other  procedures  have  not  yet  been  updated  to  the  new  methodology  and  use  only  the  ASA  classification.        GHS  Goal  –  We  want  to  have  a  Standardized  Infection  Ratio  (SIR)  no  different  or  less  than  1.0  for  each  surgical  procedure  we  monitor.    This  is  indicated  by  the  confidence  interval  crossing  1.0  (no  different  than  expected)  or  lying  completely  below  1.0  (statistically  better  than  expected).        GHS  Results  –  Data  is  reported  for  2010,  2011  and  2012.    

Page 20: January 2013 Quality and Safety Narrative Report

  20  

In  2012,  the  overall  surgical  site  infection  SIR  across  all  facilities  is  statistically  better  than  expected  at  0.65.    Four  surgery  types,  coronary  artery  bypass  grafting  (CABG),    small  bowel  resection,  colon  resection  and  Caesarian  section  all  also  have  SIRs  that  are  statistically  better  than  expected.      One  surgery  type,  hip  replacement  had  a  SIR  that  is  statistically  worse  than  expected.    The  campus  that  is  most  associated  with  this  higher  than  expected  SIR  is  the  GMH  campus.    While  PMH  hip  SIR  is  no  different  than  expected,  this  campus  has  seen  a  concerning  increase  in  orthopedic  SSIs  (  hip  and  knee  replacements).    Both  campuses  (GMH  and  PMH)  have  conducted  investigations,    which  have  resulted  in    changes  to  reduce  hip  and  in  the  case  of  PMH,  also  knee  infection  risks.    Areas  of  focus  include,  but  are  not  limited  to  ,  surgical  skin  preparation,  DVT  prophylaxis    (to    reduce    wound  hematoma  development,  a  risk  factor  for  SSI),  and  the  OR  environment.      All  other  surgery  types  have  SIRS  that  were  not  statistically  different  from  the  expected  number  of  infections.      These  surgery  types  were  bariatric  surgery,  abdominal  hysterectomy,  ventral  hernia  repair  and    knee  replacement.      

   

Central  L ine-­‐Associated  Bloodstream  Infections  (CLABSI)    Background  –  Historically,  CLABSI  rates  at  GMH  and  NG  LTACH  have  been  significantly  higher  than  the  NHSN  mean  on  many  units.    During  the  past  five  years,  the  CLABSI  Elimination  team  under  the  leadership  of  Dr.  Bill  Curran  has  focused  on  the  implementation  of  evidence-­‐based  interventions  including  a  central  line  insertion  check-­‐list  and  more  recently  a  maintenance  bundle.      GHS  Goal(s )  –  We  have  historically  set  our  goals  at  the  NHSN  mean.    This  goal  was  surpassed  during  2011    and  thus  the  target  has  been  increased  to  top  quartile.    Ultimately,  the  goal  is  to  eliminate  all  CLABSI  infections.    The  NHSN  mean  for  the  Adult  CLABSI  is  1.23/1000  central  line  days  and  the  top  quartile  is  0.22/1000  central  line  days.    Pediatric  ICU  (PICU)  pooled  mean  is  2.2/1000  central  line  days  and  the  top  quartile  is  0/1000  central  line  days.    Pediatric  Medical/  Surgical  unit  pooled  mean  is  1.50/1000  central  line  days  and  the  top  quartile  is  0/1000  central  line  days.    GHS  Results  –  The  data  in  this  report  is  presented  as  quarterly  CLABSI  rates  for  GHS  wide  adult  care,  GMH  ICU  and  Non-­‐ICU,  North  Greenville  and  PICU  areas.  The  collaborative  efforts  of  the  ICU,  Non-­‐ICU    and  North  Greenville  LTACH  CLABSI  Elimination  teams  led  to  a  continued  reduction  in  the  GHS  Adult  CLABSI  rate  to  0.32/1000  central  line  days  during  the  fourth  quarter  of    2012.  This  rate  is  less  than  the  NHSN  pooled  mean  of  1.2  /1000  central  line  days  and  very  close  to  the  top  quartile  rate  of  0.22/1000  central  line  days.    During  2012,  it  is  estimated  that  we  prevented  127  adult  CLABSIs  from  our  baseline  in  2009.    This  projects  to  an  estimated  13  to  25  lives  saved  and  an  estimated  cost  avoidance  of  approximately  5  million  dollars  ($40,000/case)  in  2012.    The  work  of  the  CLABSI  teams  was  recognized  during  the  October  2012  LDR  meeting  with  a  Quality  Pillar  award.    The  GMH  adult  CLABSI  rate  continues  to  decline  as  reflected  by  a  2012  fourth  quarter  rate  of  0.38/1000  which  is  the  lowest  it  has  ever  been.    The  2012  Pediatric  Intensive  Care  Unit  (PICU)  CLABSI  rate  is  1.5/  1000  central  line  days  which  is  less  than  the  NHSN  mean.    Pediatric  Medical/  Surgical  units    experienced  

Page 21: January 2013 Quality and Safety Narrative Report

  21  

a  2012  CLABSI  rate  of  2.1/1000  central  line  days  which  is  higher  than  the    NHSN  mean  and  no  different  from  the  previous  year.    North  Greenville  experienced  a  2012  CLABSI  rate  of  0.40/1000  central  line  days  which  is  significantly  lower  than  the  2011  CLABSI  rate  of  1.42/1000  central  line  days.    During  2012  Greer,  Patewood  and  Hillcrest  have  not  experienced  a  CLABSI.        

Venti lator-­‐Associated  Pneumonia  (VAP)    now  Venti lator-­‐Associated  Event  (  VAE)    

Background  –The  VAP  Process  Improvement  Team  led  by  Dr.  Armin  Meyer,  modified  the  oral  hygiene  procedure  to  include  the  use  of  Chlorhexidine  Gluconate  (CHG).    The  expanded  use  of  the  CASS  tube  (continuous  aspiration  of  subglottic  secretions),  head  of  bed  elevation  focus  and  extensive  education  has  led  to  significant  VAP  rate  improvement.    The  data  presented  is  in  terms  of  actual  infections  per  1000  ventilator  days.    During  2012,  the  CDC  modified  the    adult  VAP  definition  to  facilitate  a  more  standardized  methodology  that  could  be  interpreted    across  hospitals.    This  revision  has  led  to  a  definition  that  is  now    called  a  Ventilator  Associated  Event  (VAE).    GHS  transitioned  to  the    adult  VAE  definition  during  October  2012.  This  definition  change  will  be  noted  on  the  graphics  and  the    number  of  cases  can  be  impacted  by  the  change  in  definition.    Pediatrics    will  continue  to  use  the  existing    VAP  definition.    Pediatrics  has  a  quality  team  that  reviews    and  analyzes  data.        GHS  Goal  –  The  2013  GHS  goal  is  to  strive  for  top  quartile  performance  in  preventing  VAEs.    As  the  surveillance  definition  has  changed,  the  year    will  also  be  about  establishing  a  baseline  with  the  new    VAE  definition  for  adults.    The  PICU  has  a  VAP  goal  of  zero.      GHS  Results  –  GMH  and    NG  LTACH    VAP/  VAE    rates  are  presented  by  2011  calendar  year  followed  by  2012  quarterly  rates.          GHS-­‐wide,  the  adult  VAP/  VAE    rate  for  2012  is  1.17/1000  ventilator  days  which  is  significantly  lower  than  the  2011  VAP  rate  of    2.07/1000  ventilator  days.    92  VAP  have  been  prevented  in  2012  relative  to  our  baseline  in  2007,  projecting  to  an  estimate  of  18-­‐28  lives  saved  and    a  cost  savings  of  $3.68  million.        The  2012  GMH  ICU  VAP/  VAE    rate  was  1.2/1000  ventilator  days  which  is  significantly  lower  than  the  previous  years  rate  of  2.3/1000  ventilators  days.        The  PICU  has  gone  1,366  days  without  a  VAP.    (From  May  5,  2009  to  December  31,  2012).        Greer  Memorial,  Hillcrest  Memorial  and  Patewood  Memorial  have  a  VAP  rate  of  0.        North  Greenville  Long  term  Acute  Care  VAP  rate  has  steadily  declined  over  the  past  several  years  and  this  trend  continued  during  2012.    The  VAP/  VAE  rate  for  2012  was  0.89/1000  ventilator  days  and  is  

Page 22: January 2013 Quality and Safety Narrative Report

  22  

lower  than  the  2011  VAP  rate  of  1.32/1000  ventilator  days.      The  NHSN  mean  for  this  patient  population  is  o.6/  1000  ventilator  days.  

   

Catheter-­‐  associated  Urinary  tract   Infections  (CAUTI)    

Background  –      Approximately  40%  of  healthcare  associated  infections  are  Catheter  Associated  Urinary  Tract  Infections  (CAUTI)  and  the  infection  is  considered  a  hospital  acquired  condition  by  the  CMS.    GHS  has  chosen  to  aggressively  address  these  infections  with  the  goal  to  reduce  CAUTI  incidence.      Historically,  the  GMH  campus  has  not  conducted  CAUTI  surveillance  facility-­‐wide.    Key  to  prevention  is  to  identify  the  volume  of  infections  in  order  to  identify  areas  of  opportunity.      During  2011  on  the  GMH  campus,  CAUTI  surveillance  was  conducted  in  the  ICU.    With  the  addition  of  an  infection  preventionist  during  the  first  quarter  of  2012,  surveillance  was  spread  to  all  nursing  units  in  April,  2012.  Thus,  the  GMH  campus  is  still  in  the  process  of  collecting  baseline  data  which  takes  a  year.    It  is  clear  from  the  data  already  collected,  that  CAUTI  is  an  area  of  opportunity  for  the  organization.    North  Greenville,  Greer,  Patewood  and  Hillcrest  have  been  conducting  facility  wide  CAUTI  for  a  longer  period  of  time.        GHS  Goal  –  The  goal    is  to  strive  to  reduce  CAUTI  rates  to  the  top  quartile  benchmark.    The  2013  interim  goal  will  be    to  reduce  CAUTI  rates  to  the  NHSN  pooled  mean  of  2.4/1000  Foley  days  for  ICUs    and  1.5/1000  Foley  days  for  non-­‐ICUs.    Results  –  GMH  and  North  Greenville  CAUT  I  data  is  presented  by  quarterly  rates.    GMH  Non-­‐ICU  is  presented  by  monthly  rates  as  there  is  limited  data.    Greer,  Patewood  and  Hillcrest  data  is  presented  by  annual  rates.      GMH  ICU  experienced  a  2011  CAUTI  rate  of  3.94  /  1000  Foley  days  which  is  above  the  NHSN  pooled  mean  of  2.4  /1000  Foley  days.      During  2012,  the  ICU  CAUTI  rate  increased  to  5.83/1000  Foley  days.      Of  note  ,  the  NTICU,  which  was  involved  in  the  6  sigma  project  saw  a  downward  decline  in  CAUTI  during  the  end  of  the  year  as  a  result  of  performance  improvement  efforts.      GMH  Non-­‐ICU  surveillance  has  been  conducted  for  nine  months,  April  –December  ,  2012.    The  quarterly  CAUTI  rates  during  this  time  are  significantly  above  the  NHSN  pooled  mean  of  1.5.    The  GMH  PICU  CAUTI  rate    was  3.8/1000  Foley  days  for  2012,  which  is  higher  than  the  NHSN    mean  of  2.2/1000  Foley  days.    Greer    and  Patewood  experienced  2012  CAUTI    rates  below  the  NHSN  mean    Hillcrest  experienced  a  2011  CAUTI  rate  of  8.26/1000  Foley  days  and  a  CAUTI  rate  of  2.05/1000  Foley  days  during  2012.    The  reduction  in  CAUTI  at  Hillcrest  is  a  result  of  a  focus  on  catheter  maintenance.    North  Greenville  has  been  addressing  CAUTI  since  mid-­‐2011.    The  facility  initiated  a  performance  improvement  team  and  became  involved  in  the  SCHA  CAUTI  prevention  initiative.    They  focused  on  getting  catheters  out  and  on  the  maintenance  of  the  devices  while  they  are  needed.    Their  CAUTI  rate  in  2011  was  5.76/1000  Foley  days.    The  CAUTI  rate    for  2012  is    4.24/1000  Foley  days  and  reflects  a  reduction  from  the  previous  year  ,  which  remains  above  the  mean  of  2.53.    

Page 23: January 2013 Quality and Safety Narrative Report

  23  

A  Lean  Six  Sigma  project  is  being  conducted  on  the  GMH  and  North  Greenville  campuses.    Key  indentified  root    causes  include  inconsistent  compliance  with  existing  policy  for  insertion  and  maintenance  of  Foley  catheters.    Examples  of  analysis  conducted  during  the  Lean  Six  Sigma  process  are  depicted  in  some  of  the  PowerPoint  slides.    An  education  plan  was  initiated  and  is  being  finalized  during  January  2013  which  will  address  process  and  cultural  issues.    Two  CAUTI  performance  improvement  teams  are  to  be  established  to  facilitate  the  improvement  process  and  will  be  led  by  Dr.  Azim  Surka  (ICU)  and  Dr.  Bill  Kelly  (non-­‐ICU).    

   

Mult i -­‐Drug  Resistant  Organisms    Background  –  Multi-­‐drug  resistant  organisms  are  bacteria  that  have  mutated  over  time  to  become  resistant  to  most  antibiotics.    They  primarily  include  Methicillin  Resistant  Staphylococcus  Aureus  (MRSA),  Vancomycin  Resistant  Enterococcus  (VRE),  and  Clostridium  Difficile.    Individuals  can  be  colonized  with  the  bacteria,  meaning  that  the  bacteria  are  present,  but  not  causing  an  infection.    The  bacteria  can  also  cause  very  serious,  life-­‐threatening  infections.    We  are  seeing  more  individuals  come  into  the  hospital  already  colonized  with  the  bacteria.    Generally,  it  is  very  difficult  to  get  rid  of  this  colonization.    No  national  benchmarks  for  incidence  of  new  infections  are  available.        Throughout  2009,  MRSA  PCR  (polymerase  chain  reaction)  testing  was  implemented  on  the  GMH  campus,  which  allows  us  to  rapidly  determine  patients  who  are  colonized  with  the  bacteria.    All  chronically  ill  adult  patients  admitted  to  GMH  and  NGH  are  currently  being  tested  for  MRSA.    This  screening  facilitates  the  placement  of  patients  with  MRSA  colonization  into  contact  precautions  to  prevent  transmission  to  other  patients.    Patients  with  a  history  of  MRSA  whose  PCR  screening  was  negative  are  taken  out  of  precautions.    The  impact  of  Clostridium  Difficile  (CD)  has  been  felt  across  the  entire  spectrum  of  healthcare  and  is  now  recognized  as  a  pathogen  capable  of  causing  human  suffering  to  a  degree  matching  that  of  MRSA.    It  is  for  this  reason  that  this  infection  is  being  monitoring  at  GHS.    It  should  be  noted  that  beginning  January  2013,  MRSA  bacteremia  and  C.  difficile  will  be  reported  to  CDC  NHSN  and  eventually  to  CMS.      GHS  Goal  –  Goals  have  been  established  based  on  GHS  historical  data  for  each  facility  as  there  are  no  national  benchmarks.    

 MRSA  –  This  report  focuses  on  the  GMH  and  NG  healthcare  associated  (HA)  MRSA  infection  rate  whose  goal  is  to  strive  to  maintain  a  stable  rate.        Clostridium  Difficile  –  GMH  and  North  Greenville  locations  are  the  focus  of  this  report.  The  goal  is  to  establish  a  baseline  rate  with  the  C.  difficile  PCR  test  in  use,  which  is  more  sensitive  and  will  likely  increase  the  rates.    North  Greenville’s  target  is  13.7/1000  patient  days.          

   

Page 24: January 2013 Quality and Safety Narrative Report

  24  

 Results      MRSA:  GMH  MRSA  healthcare  associated  infection  (HAI)  rate  data  is  reported  in  an  annual  rate  for    2011  and    in  quarterly  rates  for    2012.    The  MRSA  healthcare  associated  infection  rate  for  2012  is    0.4/1000  patient  days  which  is  at  the  GMMC  mean  and  suggests  stabilization.      During  2012,  8%  of  adult  patients  cultured  for  MRSA  PCR  on  admission  were  positive.    Patients  colonized  or  infected  with  MRSA  continue  to  be  placed  in  contact  precautions  as  a  control  measure.      The  2012  North  Greenville  MRSA  HAI  rate  was  0.73/1000  patient  days  and  is  slightly  higher  than  the  2011  rate  of  0.45/1000  patients  days.    All  patients  are  placed  in  contact  precautions  due  to  the  level  of  endeminicity  of  MDROs.    C.  Diff ic i le  (CD):  The  GMH  CD  rate  for  2012  was    6.7/10,000  patient  days  and  cannot  be  compared  to  the  previous  year  due  to  the  implementation  of    a  C.  Difficile  PCR  which  is  more  sensitive  and  will  identify  more  cases.    When  looking  at  the    2012  quarterly  CD  rates,  an  increase  is  seen  during  the  last  two  quarters  of  the  year  which  are  directly  associated  to  the  initiation  of  this  better  test.    For  this  reason,  a  baseline  with  the  C.  difficile  PCR  is  needed  to  establish  an  organization  rate.      North  Greenville  experienced  an  increase  in  their  CD  rate  during  2011  to  16.76/10,000  pt  days.      A  reduction  of  the  CD    infection  rate  (12.21/10,000  patient  days  )  occurred  during  2012.    Of  key  interest,  is  the  reduction  occurred  with  the  implementation  of    the  PCR  test,    which  is  more  sensitive.    In  both  facilities,  environmental  cleaning  has  been  emphasized  with  hypochlorite  and  antibiotic  utilization  is  being  monitored.      Combined  MDRO:    As  part  of  our  ongoing  research  related  to  hand  hygiene,  we  have  initiated  analyses  of  a  “combined  MDRO”  outcome.    The  combined  MDRO  data  include  MRSA  infection,  facility-­‐acquired  VRE  and  facility-­‐acquired  C.  Difficile.    In  order  to  compare  data  over  time,  we  have  included  only  the  21  inpatient  units  at  GMH  that  have  had  MDRO  surveillance  since  2009.    We  examined  both  the  rates  of  combined  MDRO  per  1000  patient-­‐  days  and  the  rates  of  “clusters”  per  100  unit-­‐months.    A  cluster  was  defined  as  a  unit  monthly  rate  that  was  statistically  higher  than  expected  (i.e.  greater  than  the  upper  limit  of  the  95%  confidence  interval  for  the  unit’s  annual  rate).        The  combined  MDRO  rate  has  decreased  from  2.38  per  1000  patients-­‐days  in  2009  to  1.86  in  2012  (Jan-­‐Sept)  and  the  combined  MDRO  clustering  rate  has  decreased  from  53.6  per  100  unit-­‐months  in  2009  to  34.9  in  2012;  these  differences  were  both    statistically  significant.  In  addition,  linear  trend  testing  indicates  that  both  MDRO  rates  and  clustering  rates  have  decreased  significantly  over  time.    Future  research  will  focus  on  potential  correlation  of  these  findings  with  increased  hand  hygiene  compliance  and  other  infection  control  initiatives.              

   

Page 25: January 2013 Quality and Safety Narrative Report

  25  

APPENDIX  A    GHS  Qual ity  Review  of  S ite-­‐Specif ic  5-­‐Year  Cancer  Survival  Rates    For  Standard  4.7  Study  of  Patient  Outcomes  Presented  and  discussed  at  Cancer  Care  Committee  Meeting.  September  25,  2012  Presented  by:  Brian  McKinley,  MD                Study  Purpose:      To  compare  site-­‐specific  5-­‐year  survival  data  from  the  GHS  Cancer  Registry  to  the  most  recent  data  available  from  the  National  Cancer  Database  (NCDB).    Methods:      The  study  included  “analytic”  cases  diagnosed  with  cancer  in  2003,  2004  and  2005  (updated  by  NCDB  August  2012).    “Analytic”  cancer  cases  are  those  who  were  diagnosed  or  received  their  first  course  of  treatment  at  GHS.  GHS  survival  rates  were  compared  to  rates  from  NCDB  participating  hospitals  (n=1474  hospitals).    Rates  were  formally  compared  for  statistical  significance  using  95%  confidence  intervals.  Estimates  of  the  cumulative  proportion  of  patients  surviving  at  5  years  were  calculated  using  the  Kaplan-­‐Meier  product-­‐limit  method.        Results  -­‐Comparison  of  Overal l  Survival  Rates:    Overall  “combined-­‐stage”  GHS    5-­‐year  survival  rates  were  comparable  to  NCDB  rates  for  all  11  cancer  sites  [See  Figure  1] .    GHS  rates  were  slightly  higher  than  NCDB  rates  for  9  of  11  cancer  sites  and  slightly  lower  for  2  of  the  11  sites;  however,  these  differences  were  not  statistically  significant.  

   Recommendations:        No  fol low-­‐up  required.    Review  with  each  NCDB  update.      

   

Page 26: January 2013 Quality and Safety Narrative Report

  26  

APPENDIX  B  DEFINITIONS  FOR  QUALITY  &  PATIENT  SAFETY  DASHBOARD  

   

Measure   Definit ion   Source   Reporting  Frequency  

CMS  Al l  Care  Measures  

Also  known  as  the  “perfect  process”  score,  the  ACM  is  a  measure  of  the  number  of  times  patients  received  all  the  care  they  were  eligible  for.    See  attachment  titled,  “All  Care  Measures.”    

Premier  Quality  Measure  Tool  

Quarterly      

Value  Based  Purchasing  

The  Deficit  Reduction  Act  of  2005  directed  the  Center  for  Medicare  and  Medicaid  Services  (CMS)  to  develop  a  Value  Based  Purchasing  incentive  program  to  begin  to  align  Medicare  payments  with  hospital  quality  performance.    During  the  initial  year  of  the  VBP  program  70%  of  the  total  performance  score  will  be  based  on  clinical  measures  and  30%  will  be  based  on  patient  satisfaction  measures.  

Premier  Quality  Measure  Tool  

Quarterly  

Mortal ity  Rate   The  rate  is  calculated  by  dividing  the  total  number  of  inpatient  deaths  by  the  total  number  of  inpatients.    It  is  a  severity-­‐adjusted  mortality  rate  that  utilizes  the  3M  APR-­‐DRG  methodology.    The  benchmark  is  provided  by  Premier  Clinical  Advisor,  a  national  repository  of  hospital  and  clinical  data.  

Premier  Clinical  Advisor  

Quarterly      

Readmission  Rate   The  Readmission  Rate  is  calculated  by  dividing  the  number  of  inpatients  that  are  readmitted  to  the  same  facility  within  30  days  of  discharge,  regardless  of  the  reason  they  were  readmitted,  by  the  total  number  of  admissions.    Consequently,  the  rate  includes  both  avoidable  and  unavoidable  readmissions.    The  readmissions  are  categorized  according  to  the  initial  hospital  admission  specialty  or  service.    The  rates,  risk  adjustments  and  comparative  benchmarks  are  calculated  using  Premier  Clinical  Advisor,  a  national  repository  of  hospital  and  clinical  data.  Patients  excluded  from  the  calculations  include  patients  readmitted  from  a  skilled  nursing  facility,  patients  with  a  diagnosis  of  false  labor  and  patients  who  are  discharged  and  readmitted  the  same  day.  

Premier  Clinical  Advisor  

Quarterly  

Culture  of  Safety  Survey  

An  overall  measure  of  the  culture  of  safety  of  the  organization.    The  survey  examines  patient  safety  culture  from  the  hospital  staff  perspective.    Clinical  staff,  non-­‐clinical  support  staff  and  medical  staff  participated  in  this  survey.    Responses  were  submitted  on-­‐line.  GHS  organizational  results  were  benchmarked  against  the  AHRQ  2008  Comparative  Database  results.  

Organizational  Survey  

Annually  

National  Patient  Safety  Goals  

A  measure  of  promotion  of  specific  improvements  in  Patient  Safety,  based  on  Joint  Commission’s  highlight  of  problematic  areas  in  healthcare.    Compliance  is  measured  by  audits  of  each  element  of  performance.    Recognizing  that  sound  system  design  is  intrinsic  to  the  delivery  of  safe,  high  quality  healthcare,  the  goals  generally  focus  on  system-­‐wide  solutions,  wherever  possible.  

Observational  and  medical  record  

audits  

Quarterly  

Page 27: January 2013 Quality and Safety Narrative Report

  27  

Event  Reporting  Rate  (per  1000  patient  days)  

A  measure  of  reporting  of  quality  and  patient  safety  events  or  incidents  as  well  as  near  misses  or  “Good  Catches.”    Total  Reported  Event  Rate  is  calculated  by  dividing  the  number  of  events  reported  by  the  number  of  patient  days  times  1000.    The  data  includes  298  event  types  separated  into  patient,  visitor,  and  unsafe  conditions  (unrelated  to  an  individual).    The  inpatient  event  rate  is  benchmarked  against  event  reporting  rates  for  comparable  organizations  associated  with  University  HealthSystem  Consortium  (UHC).  

Event  Reporting  System  

Quarterly  

Patient  Fal l  Rate(per  1000  patient  days)  

A  measure  of  quality  and  patient  safety.    Falls  generally  result,  at  least  in  part,  from  the  patient  condition  and  are  most  often  caused  by  disease  state,  weakness,  confusion  and  medications.    The  benchmark  was  established  for  comparable  organizations  associated  with  University  HealthSystem  Consortium(UHC).  

NDNQI   Quarterly  

Medication  Error  Rate(per  1000  patient  days)  

A  measure  of  quality  and  patient  safety.    The  rate  of  error  reflects  variation  in  the  systems  or  processes  of  physician  ordering,  transcription,  pharmacy  dispensing,  and  nursing  administration  of  medications.    The  rate  is  calculated  by  dividing  the  number  of  events  reported  by  the  number  of  patient  days  times  1000.    An  error  is  defined  as  the  wrong  drug,  dose,  route,  time,  or  patient.    Although  error  is  inherent  in  all  human  processes,  the  benchmark  should  reflect  a  goal  as  close  to  0%  as  possible.    There  are  no  national  standard  benchmarks  for  medication  error  rates.  

Event  Reporting  System  

Quarterly  

Reported  Events  with  Harm  (per  1000  

patient  days)  

A  measure  of  harm  rate  of  reported  events.    The  harm  rate  is  calculated  by  dividing  the  number  of  events  with  harm  by  the  number  of  patient  days  times  1000.    All  event  types  are  included  in  this  rate.  Harm  scores  are  grouped  as  “Near  Miss”  which  includes  an  unsafe  condition,  or  a  near  miss.    A  second  grouping  is  events  that  “Reached  the  Patient”  including  those  that  involved  no  evident  harm,  emotional  distress  or  inconvenience,  and  those  requiring  additional  treatment.    A  third  grouping  is  those  “With  Harm”  including  events  with  temporary  harm  to  patients,  permanent  harm  to  patients,  severe  permanent  harm,  or  death.    Events  with  harm  are  benchmarked  against  comparable  UHC  participating  organizations.      

Event  Reporting  System  

Quarterly  

Hand  Hygiene   A  measure  of  quality  and  patient  safety.    Hand  hygiene  in  multiple  studies  has  been  shown  to  reduce  the  incidence  of  healthcare  associated  infections.    Compliance  is  determined  as  the  number  of  appropriate  hand  hygiene  activities  (i.e.,  washing  hands  or  applying  gel)  over  the  total  number  of  opportunities.  Observations  are  conducted  by  dedicated  and  trained  nursing  staff.  

Observational  Data  

Quarterly  

Surgical  S ite  Infection(SSI)  

An  infection  involving  the  surgical  site  post-­‐operatively.    SSIs  can  be  superficial,  deep,  or  within  an  organ  space  (e.g.,  a  joint  in  knee  surgery).    Cellulitis  of  an  incision  is  not  considered  an  infection.  

ICP  Surveillance   Monthly  

Central  L ine-­‐Associated  Bloodstream  

An  infection  in  the  bloodstream  that  is  associated  with  a  central  line  (an  intravascular  catheter  that  terminates  at  or  close  to  the  heart  or  in  one  of  the  great  vessels).  

ICP  Surveillance   Monthly  

Page 28: January 2013 Quality and Safety Narrative Report

  28  

Infection(CLABSI)  Venti lator-­‐Associated  

Pneumonia(VAP)  

A  diagnosis  of  pneumonia  in  a  patient  who  is  on  a  ventilator.  

ICP  Surveillance   Monthly  

Mult i-­‐drug  Resistant  Organisms(MDROs)  

Bacteria  and  other  microorganisms  that  have  developed  resistance  to  antimicrobial  drugs.  

ICP  Surveillance   Monthly  

Methici l l in  Resistant    Staph  Aureus(MRSA)  

A  type  of  Staph  that  is  resistant  to  certain  antibiotics.  MRSA  frequently  causes  infections  in  the  community  as  well  as  hospitals.    To  prevent  transmission  of  MRSA  in  hospitals,  patients  who  are  colonized  (the  germ  is  living  on  the  bodies,  but  isn’t  causing  infection)  are  placed  on  contact  precautions  as  well  as  those  patients  who  are  infected  with  MRSA.  

ICP  Surveillance   Monthly  

Clostr idium  Diff ic i le  (C.  Diff ic i le)  

A  spore-­‐forming  anaerobic  (grows  in  environment  without  oxygen)  bacteria.    The  most  serious  cause  of  antibiotic-­‐associated  diarrhea  (AAD)  and  can  lead  to  pseudomembranous  colitis,  a  severe  infection  of  the  colon,  often  resulting  from  eradication  of  the  normal  gut  flora  by  antibiotics.    The  C.  Difficile  bacteria,  which  naturally  reside  in  the  body,  becomes  overpopulated;  the  overpopulation  is  harmful  because  the  bacterium  releases  toxins  that  can  cause  bloating,  constipation,  and  diarrhea  with  abdominal  pain,  which  may  become  severe.  

ICP  Surveillance   Quarterly