Istroke patients receiving treatment, with hospital acquired pneumonitis, or dying in hospital

Stroke 1: Urgent brain scan for stroke patients

Metric

Proportion of stroke patients who have a brain scan (CT or MRI) performed on the day of admission.

Numerator

Discharges among cases meeting the inclusion and exclusion rules for the denominator with OPCS procedure codes for CT or MRI brain scan (U05, U21) and associated procedure dates the same as admission date.

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64).

Exclusions

Patients who die on the day of admission.

Data Source

SUS - CDS

Time frame

April 2009 - March 2010

Basis

Acute Trust

Statistical methods used

Unadjusted proportions will be presented (in-line with the Stroke Sentinel Audit reports)

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

Based on Australian Council on Healthcare Standards (2009); and

American Academy of Neurology, American College of Radiology, Physician Consortium for Performance Improvement, National Committee for Quality Assurance (2009) Stroke and stroke rehabilitation physician performance measurement set.

Adapted by the Dr Foster Unit at Imperial College

Stroke 2: Thombolysis of stroke patients

Metric

Proportion of stroke patients receiving thrombolysis treatment

Numerator

Discharges among cases meeting the inclusion and exclusion rules for the denominator with OPCS codes for thrombolysis (fibrinolytic drugs, X83.3)

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64) aged 18 to 80.

Exclusions

- Patients for whom thrombolysis is not licensed (under 18 and over 80 years old).

Nb. Thomrbolysis is not appropriate for haemorrhagic strokes; however, one of the codes (I64) includes both ischaemic and haemorrhagic stroke and so, to prevent bias resulting from varied use of this code, haemorrhagic codes are included

Data Source

SUS - CDS

Time frame

April 2009 - March 2010

Basis

Acute Trust

Statistical methods used

Unadjusted proportions will be presented

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

Based on AHRQ PSI indicators

www.qualityindicators.ahrq.gov/psi_overview.htm

Translated by the Dr Foster Unit at Imperial College

Stroke 3: Hospital acquired pneumonia due to swallowing problems

Metric

Proportion of stroke admissions hospital acquired pneumonitis due to food or vomiting

Numerator

Discharges among cases meeting the inclusion and exclusion rules for the denominator with ICD10 codes for pneumonia as follows:

• J690: Pneumonitis due to food and vomit

where the episode with this code has an end date at least 2 days after the admission date.

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64).

Exclusions

  • Patients who die within 2 days of admission.
  • Patients admitted with a primary diagnosis of pneumonia

Data Source

SUS - CDS

Time frame

April 2009 - March 2010

Basis

Acute Trust

Statistical methods used

Case-mix adjusted using a logistic regression model.

Logistic regression

The ratio is calculated by dividing the actual number of events by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has an RR of 100, that means that the number of events is exactly as it would be expected taking into account the standardisation factors. An RR above 100 means there were more events than would be expected; one below 100 means that fewer than expected events.

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

Based on Laudicella and Street, University of York cited in S. Leatherman, K. Sutherland, M. Airoldi (2008) Bridging the quality gap: Stroke. Quest for Quality and Improved Performance. Available at: http://www.health.org.uk/publications/research_reports/quality_gap.html

Adapted by the Dr Foster Unit at Imperial College

Stroke 4: Stroke hospital standardised mortality ratio (HSMR)

Metric

The standardised proportion of stroke patients who die in-hospital within 30 days of admission

Numerator

- Discharges among cases meeting the inclusion and exclusion rules for the denominator , where the patient dies in hospital and after discharge between 0-29 days (inclusive) of admission.

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64).

Data Source

SUS - CDS

Time frame

April 2009 - March 2010

Basis

Acute Trust

Statistical methods used

Case-mix adjusted using a logistic regression model.

Logistic regression

The ratio is calculated by dividing the actual number of events by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has an RR of 100, that means that the number of events is exactly as it would be expected taking into account the standardisation factors. An RR above 100 means there were more events than would be expected; one below 100 means that fewer than expected events.

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

Based on National Centre for Health Outcome Development. (2009) Mortality from stroke. Available at: www.nchod.nhs.uk

Adapted by the Dr Foster Unit at Imperial College

Stroke 5: Return to normal place of residence

Metric

Proportion of stroke patients returning to their usual place of residence following hospital treatment within 56 days

Numerator

Discharges among cases meeting the inclusion and exclusion rules for the denominator with admission source and discharge destination suggesting return to usual place of residence within 56 days.

i.e.

Discharge date - Admission date = 0-55 days inclusive

AND

(DISDEST=”19”)

Or (ADMISORC=”29” And DISDEST=”29”)

Or (ADMISORC=”30” And DISDEST=”30”)

Or ((ADMISORC=”37” Or ADMISORC=”38” Or ADMISORC=”39”) And (DISDEST=”37” Or DISDEST=”38” Or DISDEST=”39”))

Or (ADMISORC=”48” And DISDEST=”48”)

Or (ADMISORC=”50” And DISDEST=”50”)

Or (ADMISORC=”54” And DISDEST=”54”)

Or ((ADMISORC=”65” Or ADMISORC=”66” Or ADMISORC=”69”) And (<DISDEST=”65” Or DISDEST=”66” Or DISDEST=”69”))

Or (ADMISORC=”84” And DISDEST=”84”)

Or (ADMISORC=”85” And DISDEST=”85”)

Or (ADMISORC=”86” And DISDEST=”86”)

Or (ADMISORC=”88” And DISDEST=”88”)

Or (ADMISORC=”89” And DISDEST=”89”)

Or (ADMISORC=”89” And (DISDEST=”85” Or DISDEST=”86” Or DISDEST=”88”))

Or ((ADMISORC=”85” Or ADMISORC=”86” Or ADMISORC=”88”) And DISDEST = “89”).

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64).

Exclusions

• Admissions which end in death

Data Source

SUS - CDS

Time frame

April 2009 - March 2010

Basis

Acute Trust

Statistical methods used

Case-mix adjusted using a logistic regression model.

Logistic regression

The ratio is calculated by dividing the actual number of events by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has an RR of 100, that means that the number of events is exactly as it would be expected taking into account the standardisation factors. An RR above 100 means there were more events than would be expected; one below 100 means that fewer than expected events.

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

National Centre for Health Outcome Development. Returning to usual place of residence following hospital treatment: Stroke. Available at: http://nchod.nhs.uk

Adapted by the Dr Foster Unit at Imperial College

Stroke 6: Emergency readmission to hospital following treatment from a stroke

Metric

Percentage of patients of all ages with emergency readmission to any hospital (for any reason) within 27 days (inclusive) of the last, previous discharge from hospital after admission with a stroke

Numerator

Discharges among cases meeting the inclusion and exclusion rules for the denominator with an emergency admission within 0-27 days (inclusive)

Denominator

All admissions with an episode with a primary diagnosis of stroke (I60-64).

Exclusions

• Admissions which end in death

Data Source

SUS - CDS

Time frame

April 2009 - March 2010 (for index admission)

Basis

Acute Trust

Statistical methods used

Case-mix adjusted using a logistic regression model.

Logistic regression

The ratio is calculated by dividing the actual number of events by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has an RR of 100, that means that the number of events is exactly as it would be expected taking into account the standardisation factors. An RR above 100 means there were more events than would be expected; one below 100 means that fewer than expected events.

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ - that is, where the trust’s rate diverges significantly from the national rate.

Notes

Based on EU Public Health Outcome Research and Indicator Collection. EUPHORIC Project. Deliverable N. 6 Detailed sheets of the collected outcome indicator (long list). February 2008. Available at: http://www.euphoric-project.eu/files/Deliverable_6_54%20indicators_0.pdf

Adapted by the Dr Foster Unit at Imperial College

Your feedback

Please share any concerns or suggestions for improvement that you might have regarding this indicator. In particular, please consider these questions:

  • Are there any diagnosis or procedure codes that have been included that you believe should be removed? Please give your reasons
  • Are there any diagnosis or procedure codes that have been omitted that you believe should be included? Please give your reasons
  • What are the strengths and weaknesses of this metric as an indicator

You can use the feedback box below to submit comments to HSJ. Alternatively, you can email Dr Foster directly at HGconsult2010@drfoster.co.uk. Please submit your response by 31 August 2010.