I02043 Contextual Indicator: Reasons for emergency readmission - specification v1.1

This indicator measures the counts and crude rates per 100,000 discharges of different primary diagnoses reported in the emergency readmission episode following a discharge that involved a specified diagnosis or procedure.  Figures are reported at a national level.

The sample is based on the same criteria as described in the annual emergency readmissions within 30 days of discharge from hospital by diagnosis (fractured proximal femur and stroke) and procedure (hip replacement surgery and hysterectomy) indicators (I02041 and I02042 respectively).

This indicator adds context to these indicators and is designed to increase the understanding of the relationship between a hospital stay and a subsequent emergency readmission.

Results are presented for males, females and persons for each discharge diagnosis or procedure group.

This indicator forms part of the Hospital Care Domain.

The records for the denominator and the numerator are taken from Hospital Episode Statistics for Admitted Patient Care (HES APC) linked as Continuous Inpatient (CIP) Spells.

This section sets out how this indicator is constructed.

Introduction

This indicator measures the counts and crude rates per 100,000 discharges of different primary diagnoses reported in the emergency admission episode following a discharge that involved a specified diagnosis or procedure.  Figures are reported at a national level.

The indicator is presented for the same breakdowns as the associated emergency readmissions indicators:

• For emergency readmissions within 30 days of discharge from hospital by diagnosis (I02041) the diagnosis at the original discharge is matched to the diagnosis at emergency readmission.  Diagnoses at discharge are presented separately for fractured proximal femur and stroke.
• For emergency readmissions within 30 days of discharge from hospital by procedure (I02042) the primary operation before the original discharge is matched to the diagnosis at emergency readmission.  Operation prior to discharge are presented separately for either hip replacement surgery or hysterectomy.

Fields and Filters

Please refer to the related indicator specifications for filters on the source dataset as this indicator uses the dataset created in the production of the associated emergency readmissions indicator.

Numerator (count of each event)

The following fields and values are used for the numerator:

The original discharge spell where the reference diagnosis or procedure took place:

DISDATE (date of discharge)

The date of discharge of the original attendance in hospital where the primary diagnosis or procedure occurred is within the financial year being reported, 1st April to 31st March inclusive.  Note: the three-year pooled results are not used.

PROCODET (Hospital Provider Code)

Valid value.  Used to aggregate counts to national level.

SEX (sex of patient)

Equals 1 (male) or 2 (female).  Note: values are used to filter into the reporting groups: male only, female only and persons (male and female).

The emergency readmission spell has:

DIAG_4_01 (primary diagnosis)

Is a valid ICD-10 code.

If the code is invalid or there is no emergency readmission then the count is categorised as “Total Discharges, with no emergency readmission”.

Denomator

The total numerator values (as calculated above) are summed for each breakdown of reporting (or discharge) year, original diagnosis or procedure and sex group.

The dataset described above is aggregated to produce counts of primary diagnosis from the first episode of the readmission spell grouped by reporting year, sex category, the diagnosis or procedure group (fractured femur, stroke, hip replacement or hysterectomy) presented by the parent indicators.

The crude rate is then calculated:

\(\mathsf{\Large \hat p = { {x} \over {n} } { \times 100,000 } }\)

where:

\(\mathsf{\hat p}\) = crude rate per 100,000 (sample proportion)

\(\mathsf{x}\) = count of specific outcomes following a discharge for the current breakdown.  Includes diagnosis on readmission and no readmission.

\(\mathsf{n}\) = sample size i.e. the sum of all discharges for the specified breakdown.

The lower and upper limits of the 95% confidence interval for the crude rate are then calculated using:

\(\mathsf{\Large \hat p \pm z \times \sqrt \frac {\hat p(1-\hat p)}{n} \times 100,000}\)

where:

\(\mathsf{\hat p} \) = sample proportion (the crude rate calculated above)

\(\mathsf{z} \) = z value from standard normal distribution for the 95% confidence level (1.96)

\(\mathsf{n} \) = sample size.

When output to Excel the following data columns are produced:

When output to csv the following data columns are produced: