OOPS Adverse Events in the Hospitalized Patient:

1
OOPS! Adverse Events in the Hospitalized Patient

• Why are they happening, and what can we do about
  them?

2
Case Presentation

• CJ, 72 yo AAF with PMHx sig for
• CAD (S/P CABG)
• AAA
• CVA x2
• DVT
• DIC
• Pancytopenia
• Presents to ED with hematemesis

3
Case Presentation

• Pt had episode of hematemesis and BRBPR prompting
  EMS activation and ED visit
• EMS unable to establish IV
• Admission vitals
• BP 173/31
• HR 132
• RR 20

4
Case Presentation

• Pt thought to be unstable by ED personnel
• Peripheral IV access attempts (including external
  jugular) unsuccessful
• PICC line inadequate for volume resuscitation
• Central line placement attempted

5
Case Presentation

• Right IJ
• Right femoral
• Left IJ
• Left subclavian
• Subclavian (? L vs. R)
• Left femoral

• Unsuccessful
• Unsuccessful
• Unsuccessful
• Unsuccessful
• Unsuccessful
• Successful

6
Case Presentation

• During line placements, pt noted to be
  hypotensive, and total volume received includes
• 6 liters normal saline
• 8 units PRBC
• Pt begins to become less responsive with
  increasing oxygenation requirements
• Decision is made to intubate

7
Case Presentation

• Intubated successfully, and CXR obtained
• CXR reveals right tension pneumothorax
• Chest tube placed on right, CXR obtained
• CXR reveals left pneumothorax and free air under
  the diaphragm
• Chest tube placed on left
• CXR reveals pneumothoraces and free air resolved

8
Case Presentation

• Pt transferred to MICU
• Free air under diaphragm thought by surgery to be
  actually behind the diaphragm, secondary to CT
  placement
• Hgb in MICU 18.6 (up from 8.3 in ED)
• Pt continued to show persistent lactic acidosis

9
Case Presentation

• EGD revealed Mallory Weiss tear
• Mesenteric angiogram negative for ischemia, but
  noted pneumoperitoneum
• Surgery requested emergent trip to OR
• Family decided against operative management, and
  shifted to comfort care
• Pt died after extubation and discontinuation of
  pressors

10
INTRODUCTION

• Clinical questions
• How common are adverse events in hospitalized
  patients?
• What is the nature of these adverse events?
• How many of them are preventable?
• How many of them are caused by negligence?
• What can we do to prevent as many of them as
  possible?

11
Medical Insurance Feasibility Study

• Retrospective chart review 1974
• 20,864 charts from 23 California hospitals
• 3,011,000 admissions in CA that year
• Results similar to data sample of 759,223
  patients from CA Health Data Corp
• Therefore results able to be generalized

12
Medical Insurance Feasibility Study

• Definitions
• Potentially compensable event
• Disability caused by health care management
• Disability
• Temporary or permanent loss of physical or mental
  function or
• Economic loss in absence of impairment
• Causation
• Disability is more probably than not attributable
  to health care management

13
Medical Insurance Feasibility Study

• Definitions (cont.)
• Health care management
• Actions (commission) or inactions (omission)
• Any health care provider or attendant
• Whether or not the action or inaction constitutes
  legal fault

14
Medical Insurance Feasibility Study

• PCE
• Lengthened hospital stay or continued after D/C
• Necessitated hospital stay
• Treated during hospital stay (even if occurred
  prior to admission)
• Present during hospital stay but not reason for
  admission and not treated, and disability 3.4
• Caused by event during stay but discovered on
  readmission

15
Medical Insurance Feasibility Study

• Disability
• 3.0 minor temporary
• lt 30 days, no surgery needed
• 3.1 minor temporary
• lt 30 days, requiring surgery
• 3.2 major temporary
• gt 30 days but lt 2 years

16
Medical Insurance Feasibility Study

• 3.3 minor permanent partial
• Permanent conditions not functionally disabling
• Loss of spleen
• Loss of uterus
• 3.4 major permanent partial
• Substantial damage but not complete loss of
  ability to perform most ordinary functions
• 3.5 major permanent total
• Substantial damage, usually sufficient to alter
  life-style to dependent position

17
Medical Insurance Feasibility Study

• 3.6 grave permanent total
• Complete dependency or short-term fatal prognosis
• 3.7 death

18
Medical Insurance Feasibility Study

• Methods
• Initial screen by medical auditors eliminated 50
  of charts but lt 1 of PCEs
• Remaining charts reviewed by one of investigators
• If PCE thought to be found, it was discussed by
  whole committee
• Rated for level disability

19
Medical Insurance Feasibility Study

• Results
• 970 PCEs found from 20,864 charts (4.65)
• Statistically significant difference in PCE rate
  by age (plt0.05)
• lt 65 years old - 4.07 /- 0.30
• gt 65 years old - 7.22 /- 0.82
• Statistically significant difference in PCE rate
  by payee (plt0.05)
• Government 6.23 /- 0.56
• Nongovernment 3.83 /- 0.36

20
Medical Insurance Feasibility Study

• Disability from PCE
• Temporary (3.0-3.2)
• Minor perm (3.3)
• Major perm (3.4-3.6)
• Death (3.7)

• Percentage
• 80
• 6.5
• 3.8
• 9.7

21
Medical Insurance Feasibility Study

• Causal factor of PCE
• Specific procedures
• Drugs and biologics
• Medical devices
• Nondiagnosis
• General medical management
• Misdiagnosis
• Anesthetic
• Nursing

• Percentage
• 66.1
• 18.8
• 4.1
• 2.7
• 2.6
• 2.2
• 2.0
• 1.6

22
Medical Insurance Feasibility Study

• Results generalized to all California admissions
  for 1974
• 4.65 PCE rate in 3,011,000 admissions
• 140,000 /- 8,700 PCEs in CA in 1974
• 112,000 are temporary (3.0-3.2)
• 9,100 are minor permanent (3.3)
• 5,300 are major permanent (3.4-3.6)
• 13,600 deaths (3.7)

23
Medical Insurance Feasibility Study

• Conclusion
• Since CA had 10 of US population, if results
  from California can be generalized to all of US
• 53,000 major permanent disabilities and 136,000
  deaths caused by hospitalization in US in 1974

24
Medical Insurance Feasibility Study

• Conclusions (cont.)
• The elderly are at almost 2x the risk for PCEs
  while in the hospital
• Patients who have government payees have 1.5x
  risk for PCEs while in the hospital

25
Harvard Medical Practice Study

• Retrospective chart review of 31,429 hospital
  records in NY from 1984
• 2,671,863 (non-?) admission in NY in 1984
• Charts reviewed for evidence of adverse events,
  presence of negligence, and degree of disability

26
Harvard Medical Practice Study
27
Harvard Medical Practice Study

• Methods
• Sample of 31,429
• Screened by nurses and medical records analyst
• If positive reviewed by 2 physicians

28
Harvard Medical Practice Study

• Physician reviewers look for
• Evidence of adverse events
• Grade of confidence (0-6) that adverse event
  occurred
• If adverse event occurred, degree of disability
  assessed
• Grade of confidence (0-6) that negligence
  occurred

29
Harvard Medical Practice Study

• Physicians could consult specialists in making
  their assessments
• Discrepancies between physician reviewers were
  resolved by supervising physician
• 1 of all charts were independently rereviewed by
  supervisor to assess reliability of initial
  physician review

30
Harvard Medical Practice Study

• Several months after initial review, missing
  records located and reviewed
• Rates of adverse events and negligence assessed
  in these records to estimate rates for still
  missing records
• Presence of adverse event or negligence required
  at least 4/6 confidence
• Disability scores averaged

31
Harvard Medical Practice Study

• Disability
• Minimal impairment, recovery 1 month
• Moderate impairment, recovery 1-6 months
• Moderate impairment, recovery gt 6 months
• Permanent impairment, lt 50 disability
• Permanent impairment, gt 50 disability
• Death

32
Harvard Medical Practice Study

• Results
• Incidence rate of adverse events
• 3.7 (95 CI 3.2 - 4.2) - 1133 events
• Incidence rate of negligence
• 1.0 (95 CI 0.8 - 1.2) - 280 events
• Percentage of adverse events due to negligence
• 27.6 (95 CI 22.5 - 32.6)

33
Harvard Medical Practice Study

• Based upon a total of 2,671,863 admissions in
  1984, the above data yields
• 98,609 adverse events
• 27,179 adverse events secondary to negligence
• If applicable to entire country, based on NY
  population, this yields
• 1,281,917 adverse events, of which,
• 353,327 are due to negligence

34
Harvard Medical Practice Study

• Of these 1,281,917 adverse events
• 31,691 will cause permanent impairment with gt
  50 disability
• 165,772 will cause death
• Since it may be problematic to generalize to the
  entire country, here is the breakdown for all NY
  patients

35
Harvard Medical Practice Study
36
Harvard Medical Practice Study

• The worse the injury, the more likely the adverse
  event was cause by negligence
• These differences in percentages of negligence
  were significant (plt0.0001)
• What other demographic patterns were noted?

37
Harvard Medical Practice Study

• Rates of adverse events increased strongly with
  age (plt0.0001)
• Rates of adverse events caused by negligence
  increased after age 64 (plt0.01)
• Both of these differences held even after
  standardization for DRG class

38
Harvard Medical Practice Study
39
Harvard Medical Practice Study

• Rates of adverse events varied strongly with
  specialty (plt0.0001)
• Vascular surgery 16.1
• General medicine 3.6
• Neonatology 0.6
• Rates of negligence did not vary with specialty
  (p 0.64)

40
Harvard Medical Practice Study II

• With data now revealing the quantity of adverse
  events, the next question is
• What is the nature of these adverse events?
• 599/1133 adverse events were surgical, the
  remaining events are described below

41
Harvard Medical Practice Study II
42
Harvard Medical Practice Study II

• The most common cause of nonoperative AE was an
  adverse drug reaction
• The most common causes of adverse drug reactions
  are listed below, followed by the most common
  types of complications

43
Harvard Medical Practice Study II

• Antibiotic
• Antitumor
• Anticoagulant
• Cardiovascular
• Antiseizure
• Diabetes
• Antihypertensive
• Analgesic

• 16.2
• 15.5
• 11.2
• 8.5
• 8.1
• 5.5
• 5.0
• 3.5

44
Harvard Medical Practice Study II

• Marrow suppression
• Bleeding
• CNS
• Allergic/Cutaneous
• Metabolic
• Cardiac
• GI
• Renal
• Respiratory

• 16.3
• 14.6
• 14.6
• 14.0
• 10.1
• 9.6
• 7.9
• 6.7
• 2.8

45
Harvard Medical Practice Study II

• With detail about what kind of adverse events
  take place, the following questions arise
• Where are these adverse events happening?
• Are certain populations more prone to certain
  types of event?

46
Harvard Medical Practice Study II
47
Harvard Medical Practice Study II
48
Harvard Medical Practice Study II

• With data on location and demographics of adverse
  events, the final question arises
• How is negligence associated with disability?
• Gravity of negligence was correlated to degree of
  disability as shown

49
Harvard Medical Practice Study II
50
Harvard Medical Practice Study II

• So, gravity of negligence is associated with
  increased levels of disability and death.
• How does the presence or absence of negligence
  relate to malpractice claims?

51
Harvard Medical Practice Study III

• 30,195 discharges - 1,133 adverse events
• 280 caused by negligence (1 discharges)
• 98 patients filed claims
• 47 of these were more likely than not related
  to hospitalization during study
• Of the remaining 51, 44 were definite nonmatches,
  other 7 with insufficient data

52
Harvard Medical Practice Study III

• Of the 47 malpractice claims filed
• 12 met none of the screening criteria for an AE
  and were not reviewed by MD
• Of the remaining 35 cases
• 14 no possibility of AE
• 3 less than likely AE
• 10 AE, not caused by negligence
• 8 AE caused by negligence

53
Harvard Medical Practice Study III

• When results extrapolated to NY state
• 98 of AEs due to negligence did not result in
  malpractice claims
• Of AEs due to negligence, 53 had strong evidence
  of negligence
• Of these AEs with strong evidence, 42 had
  disability gt6 months

54
Harvard Medical Practice Study III
55
Harvard Medical Practice Study III

• Overall, the frequency of malpractice claims
  statewide is 0.13 of all discharges
• The frequency of claims in cases with AEs caused
  by negligence is 1.53
• Statewide there were an estimated 3570 claims for
  27,179 AEs due to negligence.
• This estimate yields a ratio of 7.6 AEs due to
  negligence for each malpractice claim

56
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims

• 51 malpractice claims identified (47 mentioned
  earlier and 4 found on further follow-up)
• 46 of these claims were closed by 12/31/95
• 1/46 led to a jury trial, remaining 45 settled

57
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims
58
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims
59
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims

• The presence of an AE (negligent or not) was NOT
  associated with the outcome of litigation
• The presence of an AE (negligent or not) was NOT
  associated with payment to plaintiff
• In fact, there was only one predictor of payment
  in multivariate analysis

60
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims
61
Relation Between Negligent Adverse Events And The
Outcomes Of Medical-Malpractice Claims

• Conclusions
• Approximately 40 of malpractice claims are
  settled in favor of the plaintiff
• Average settlement was approximately 40,000
• Adverse events and negligence were not associated
  with outcome or payment
• Permanent disability was the only predictor of
  payment

62
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Based upon previous data regarding the rates of
  adverse events in hospitalized patients, the
  authors attempted to determine if housestaff
  cross-coverage contributed to these adverse
  events

63
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Methods
• In a previously validated study design, residents
  self-reported adverse events that occurred in
  their patients
• These cases were compared to 2 controls (the
  patients in the beds on either side of the
  patient)
• Controls were required to be on the medical
  services

64
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Methods (cont.)
• Preventability was assessed by 3 independent
  physician reviewers on a scale from 1-6
• 2 out of 3 reviewers had to agree on
  preventability rating
• Reproducibility was tested by rereviewing a
  random sample of 45 adverse events

65
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Coverage during an adverse event was assigned by
  determining who was responsible for the patient
  when the incident that led to the adverse event
  occurred
• Coverage was one of the following
• Primary intern
• Intern on same team
• Intern from different team
• Night-Float Resident (HO-II or HO-III)

66
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Results
• 3146 patients admitted to medical services
  (including ICU)
• 124 adverse events identified (4)
• 54 of these events were given preventability
  scores of at least 4 (therefore they were
  preventable)

67
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Using univariate analysis, the following clinical
  factors were associated with potentially
  preventable adverse events
• HTN, Hx GIB, Chronic Liver Dz, DNR
• Higher APACHE II Score
• Paracentesis or Colonoscopy prior to event
• Cross-coverage by intern or night-float resident

68
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Using multivariate analysis, only three risk
  factors were significant, independent correlates
  of potentially preventable adverse events
• APACHE II Score OR 1.2 per point increase (95
  CI 1.1-1.4)
• Hx GIB OR 4.7 (CI 1.2-19.0)
• Cross-coverage OR 6.1 (95 CI 1.4-26.7)

69
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Using univariate analysis, the following clinical
  factors were associated with Unpreventable
  Adverse Events
• HTN, Hx GIB, Hypercholesterolemia, and previous
  cardiac catheterization
• NO association was found between cross-coverage
  and unpreventable adverse events

70
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?
71
Does Housestaff Discontinuity Of Care Increase
The Risk For Preventable Adverse Events?

• Conclusions
• Cases who had preventable adverse events were
  6.1x more likely to have been covered by someone
  other than their primary team
• Higher APACHE II scores, and history of GI
  bleeding were associated with higher risk of
  preventable adverse events
• Cross-coverage WAS NOT associated with
  unpreventable adverse events

72
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• Based on the data from the previous study, the
  housestaff at Brigham and Womens decided to
  develop a program to help eliminate these
  preventable cross-coverage adverse events
• They designed a computerized sign-out program to
  facilitate cross-coverages knowledge of the
  patient

73
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• The sign-out included
• Summary of current medical status
• Resuscitation status
• Recent laboratory values
• Allergies
• Problem List
• To Do List
• The sign-out was on all hospital computers

74
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• Data on adverse events was collected (in a
  similar fashion to the previous study) during a 2
  month pre-intervention period, and then during a
  4 month intervention period
• Data from the pre-intervention period was
  compared with the baseline period (the previous
  study) and found to be similar

75
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS
76
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• During the 4 month intervention period, there
  were 3747 admissions
• 89 patients had adverse events
• 46 adverse events were preventable
• 9 preventable adverse events occurred during
  cross-coverage

77
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• Using univariate analysis, only alcohol abuse,
  APACHE II score and prior endotracheal intubation
  were associated with a higher rate of preventable
  adverse events.
• Housestaff cross-coverage WAS NOT associated with
  a higher rate of preventable adverse events

78
USING A COMPUTERIZED SIGN-OUT PROGRAM TO IMPROVE
CONTINUITY OF INPATIENT CARE AND PREVENT ADVERSE
EVENTS

• Using multivariate analysis, only Alcohol use and
  APACHE II score were associated with higher rates
  of preventable adverse events.
• House staff cross-coverage was NOT associated
  with preventable adverse events (OR 0.4 95 CI
  0.0-4.8)

79
CONCLUSIONS

• Adverse events are common in hospitalized
  patients (3.7 of hospitalizations, estimated
  1,250,000 annually)
• Negligence is a common cause of adverse events
  (27.6 of adverse events)

80
CONCLUSIONS

• There is significant morbidity and mortality from
  these adverse events (31,000 permanently
  disabling injuries and 165,000 deaths annually)
• Rates of adverse events and negligence rose with
  age

81
CONCLUSIONS

• Drug reactions are the most common nonoperative
  adverse events
• Rates of negligence in adverse events is highest
  in
• the emergency room
• the patients room

82
CONCLUSIONS

• As the gravity of negligence increases, the
  degree of disability increases
• The rate of malpractice claims being filed is
  0.15 of hospital discharges
• Only 1.5 of adverse events with negligence lead
  to malpractice claims

83
CONCLUSIONS

• Outcomes and dollar awards of lawsuits are not
  related to adverse events or negligence
• The only predictor of payment in malpractice
  lawsuits is degree of disability
• Patients being cross-covered are 6x more likely
  to have a preventable adverse event

84
CONCLUSIONS

• Computerized sign-out sheets may be one of many
  ways to reduce these preventable adverse events

85
WRAP-UP

• If other industries (such as FAA) can regulate
  themselves to prevent adverse events, it stands
  to reason that we can too
• If as many deaths occurred in airtravel as in
  healthcare, 2 jumbo jets would have to go down
  every day of the year
• The American people are not going to stand for
  this in the long term

86
WRAP-UP

• We would do well to emulate these industries, and
  adopt quality control measures to prevent adverse
  events
• Some of these solutions are already on their way

87
ACKNOWLEDGEMENTS

• Dr Moran
• The cast of the Simpsons

88
(No Transcript)
89
(No Transcript)
90
(No Transcript)