Assessment of clinical risk and outcome measures in the out of hours hospital

1
Assessment of clinical risk and outcome measures
in the out of hours hospital
Claire Gordon Specialist Registrar Intensive
Care Medicine
Daniel Beckett Specialist Registrar Acute
Medicine
Royal Infirmary of Edinburgh, UK
2
Introduction

• There is evidence that for the acutely unwell,
  care overnight is often sub-optimal
• The Hospital at Night project is being widely
  introduced across the United Kingdom as a
  mechanism to maintain quality out of hours
  healthcare in the face of the European Working
  Time Directive

3
Introduction

• Our study aimed to
• quantify the numbers of patients at risk
  overnight,
• evaluate the use of a validated SEWS system
• assess clinical outcome prior to the introduction
  of a hospital at night policy

4
Background

• The Night Time Acute Cover Working Group of the
  Royal College of Physicians of London (2002)1
  highlighted deficiencies in quality of healthcare
  overnight
• Half of all units reduced doctors on call
  overnight
• Suggested that reduced level of medical staffing
  correlates with mortality2

5
Background

• Doctors covering on average 67 patients per
  night, others more than 200
• In one third of units there were no first year
  doctors involved in overnight care, which is a
  significant training issue
• In 30 of units there was no resident registrar
  overnight, resulting in no senior supervision of
  patient care

6
Background

• NCEPOD Who Operates When? I report in 1997
  highlighted that operations overnight (0000
  0759) were carried out by less experienced
  surgeons, with less experienced anaesthetic
  support3
• Significant improvements made by 2003 (Who
  Operates When? II) but there remains a
  significant disparity between seniority of staff
  during the day and at night3

7
Background

• There is evidence of an increase in early
  mortality (within 48 hours of hospital
  admission) of patients admitted at night4
• The overall relative risk of mortality as an
  in-patient is increased for those admitted at
  night5
• Night time discharge from ITU to the ward is
  associated with increased mortality6,7

8
Background

• Traditionally overnight NHS hospitals have relied
  on speciality specific tiers of doctors with the
  most junior staff resident and consultants
  on-call from home
• Concerns regarding excessive hours, poor
  performance and detriment to patient care

9
Background

• More recently the European Working Time Directive
  (EWTD) has acted as a major catalyst for changing
  the working pattern of doctors in training
• Stipulates a maximum 56 hour working week by
  August 2007
• 48 hour week by August 2009

10
Background

• Hospital at night piloted in 2003 in UK
• Redefine how medical cover is provided in
  hospitals during the out of hours period
• Multidisciplinary night team
• Cover is defined by competency rather than
  professional demarcation

11
Background

• Key elements of Hospital at Night
• Supervised multi-speciality handover
• Extended roles of nursing staff including limited
  prescribing
• Bleep filtering through central coordination
• Moving a significant proportion of non-urgent
  work from the night into the evening or daytime
• Reducing unnecessary duplication of work

12
The Study

• The Royal Infirmary of Edinburgh, an 870 bedded
  teaching hospital
• Hospital at Night was implemented in October 2006
  
• integral part of the redesign of NHS Lothian
  acute care services
• H_at_N prospectively to cover all 18 level 1 wards
  plus 4 high dependency areas

13
(No Transcript)
14
The Study

• Prior to the implementation of Hospital at Night,
  the hospital adhered to a traditional on-call
  policy
• First year doctors on site overnight (6 PRHOs, 2
  SHOs as first point of contact)
• Uneven distribution of workload (eg orthopaedic
  house officer covering 4 wards, vascular house
  officer just 1)
• Registrars and consultants on call from home
• SHOs available but may be in theatre or AE

15
The Study

• Our study aimed to quantify the numbers of
  patients at risk overnight, evaluate the use of a
  validated SEWS system and assess clinical outcome
• To enable accurate matching of capacity and
  demand

16
Early Warning Scores

• Patients have documented clinical deterioration
  (or new complaints) in the hours leading up to
  cardiopulmonary arrest
• 84 of patients in the preceding 8 hours8
• A failure of systems to recognise and effectively
  intervene

17
Early Warning Scores

• First developed in 1997, points are allocated
  according to derangement of physiological
  parameters9
• Elevated modified Early Warning Score (MEWS) on
  admission to medical unit associated with
  increased risk of death, or need for critical
  care10

18
Early Warning Scores

• Use of MEWS encouraged by The Royal College of
  Physicians report The Interface between Acute
  General Medicine and Critical Care in 200211
• NCEPOD report An Acute Problem in 2005
  highlighted that 27 of hospitals still did not
  use an early warning system12

19
SEWS

• Scottish Early Warning Score
• Heart rate
• Blood pressure (systolic)
• Oxygen saturation via pulse oximetry
• Respiratory rate
• Temperature
• (GCS/Urine output)
• Score of 4 necessitates medical review within 20
  minutes

20
SEWS
21
SEWS

• Admission SEWS correlates with in hospital
  mortality and length of stay
• Following the introduction of the scoring system
  in the Royal Infirmary of Edinburgh, in-patient
  mortality decreased13

22
The Study

• Observational study
• 17 nights (2000-0800)
• All 18 level 1 wards plus the Combined Assessment
  Area (CAU)
• Critical care subject to separate evaluation

23
The Study

• Incidents of clinical concern or patients
  triggering SEWS 4
• Response times
• Seniority of doctor attending patient
• Initial Scottish Early Warning Score (SEWS)
• Change in SEWS as a surrogate marker for clinical
  status
• Incident outcome by 0800

24
Results

• 136 incidents of clinical concern were recorded
• 80 within the ward arc
• 20 within the Combined Assessment Area
• SEWS recording better undertaken in the Combined
  Assessment Area
• 92 had the 5 main physiological variables
  recorded, compared with 50 on the ward arc

25
Results

• 56/136 patients scored SEWS 4
• Median response time 5 minutes (mean 23 minutes,
  range 0 280 minutes)
• 82 seen within protocol prescribed 20 minutes

26
Results
27
Results

• No significant difference between surgical wards,
  medical wards and CAU with respect to response
  times to patients with SEWS 4

28
Results

• SEWS ?4

29
Results

• SEWS 4
• Outcome at 0800
• 44/56 (79) stabilised on the ward
• 7/56 (13) transferred to critical care
• 5/56 (9) deaths
• 3 cardiac arrests
• 2 patients not for escalation of treatment

30
Results

• 80/136 patients scored SEWS lt4 but caused
  clinical concern
• Median response time 10 minutes (mean 25 minutes,
  range 0 330 minutes)

31
Results
32
Results

• Response times in CAU quickest in the hospital.
  Significantly better than
• orthopaedic (plt0.001)
• vascular (plt0.001)
• general surgery (p0.04)

33
Results

• Response times in orthopaedic wards significantly
  slower than
• general medicine (p0.003)
• elderly medicine/GI (p0.01)
• cardiology/respiratory (p0.03)
• general surgery (p0.003)
• Response times in vascular slower than
• elderly medicine/GI (p0.004)
• general medicine (p0.006)
• general surgery (p0.004)

34
Results

• SEWS lt4
• Outcome at 0800
• 69/80 (86) stabilised on the ward
• 9/80 (11) transferred to critical care
• 2/80 (3) deaths
• 2 cardiac arrests

35
Results

• Seniority of attending doctor

36
Discussion

• SEWS is well utilised and understood by nursing
  staff in CAU
• dedicated education program
• Patients admitted to CAU routinely have
  observations monitored and SEWS calculated
• Ongoing issues with calculating SEWS as part of
  routine ward observations

37
Discussion

• Patients triggering SEWS 4 tend to be seen
  sooner than those with SEWS lt4 (ns) with 82
  concordance to the 20 minute guideline
• No difference between CAU and ward arc
• No significant outcome effect seen with delay in
  review (but numbers small)

38
Discussion

• No significant differences in response time for
  patients SEWS 4 across the hospital
• Significant differences arise between various
  specialty wards for patients with SEWS lt4 but
  causing clinical concern
• Important to remember that a proportion of these
  patients are still unwell with 11 requiring
  transfer to critical care
• Only 50 had full SEWS measured

39
Discussion

• Patients triggering SEWS 4 are generally
  reviewed by more senior staff than those with
  SEWS lt4 (ns)
• The majority of clinical issues overnight are
  dealt with satisfactorily by doctors of SHO grade
  or below

40
Recommendations

• Current level of medical staffing in CAU
  appropriate to allow prompt review and
  satisfactory treatment of patients triggering
  SEWS 4
• CAU thus to remain autonomous and not be formally
  covered by H_at_N
• SEWS education program to be broadened to
  encompass ward arc

41
Current Status

• Hospital at Night commenced October 2006
• 1 Specialist Registrar, 2 SHOs, 2 FY2s and 2 FY1s
  plus 3 SNPs (Senior Nurse Practioners)
• Specialist Registrar also has supervisory
  capacity over 2 SHOs and 2 FY1s working in
  Combined Assessment

42
Future Investigation

• Repeat audit now underway to assess any changes
  in incident response times and clinical outcomes
  now Hospital at Night operational
• Also audit broadened to include critical care,
  before and after the introduction of Hospital at
  Night

43
Many Thanks

• Professor Derek Bell, Professor of Acute
  Medicine, Imperial College, London
• Dr Donald MacLeod, Consultant in Acute Medicine,
  Western General Hospital, Edinburgh
• Ruth Paterson, Practice Development Nurse,
  Western General Hospital, Edinburgh

44
References

• 1. http//www.rcplondon.ac.uk/news/statements/doc_
  OutofHours01.asp
• 2. Jarman B, Gault S, Alves B et al. Explaining
  the differences in English hospital death rates
  using routinely collected data. BMJ. 1999.
  318 1515-1520
• 3. http//www.ncepod.org.uk/pdf/2003/03full.pdf
• 4. Silbergleit R, Kronick SL, Philpott S et al.
  Quality of emergency care on the night shift.
  Acad Emerg Med. 2006. 13 325-30
• 5. Hilson SD, Rich EC, Dowd B et al. Call
  nights and patient care. J Gen Intern Med.
  1992. 7 405-410
• 6. Duke GJ, Green JV, Bredis JH. Night-shift
  discharge from intensive care units increases the
  mortality risk of ICU survivors. Anaesth
  Intensive Care. 2004. 32(5) 697-701
• 7. Tobin AE, Santamaria JD. After-hours
  discharges from intensive care are associated
  with increased mortality. 2006. Med J Aust. 184
  334-7

45
References

• 8. Schein RM, Hazday N, Pena N et al. Clinical
  antecedents to in-hospital cardiopulmonary
  arrest. Chest. 1990 98 1388-92
• 9. Morgan RJM, Williams F, Wright MM. An early
  warning scoring system for detecting developing
  critical illness. Clin Intens Care 1997 8 100
• 10. Subbe CP, Kruger M, Rutherford P et al.
  Validation of a modified Early Warning Scoring
  Score in medical admissions. QMJ. 2001 94
  521-26
• 11. Royal College of Physicians of London. The
  interface between acute general medicine and
  critical care. Report of a working party of the
  Royal College of Physicians. 2002.
• 12. NCEPOD. An acute problem? 2005
• 13. Paterson R, Macleod DC, Thetford D et al.
  Prediction of in-hospital mortality and length of
  stay using an early warning system clinical
  audit. Clin Med. 2006 6 281-3