Glossary

95% CI
Age-sex standardisation
ICD
FEV1
SMR
Use of log scales
Hospital admission data (HIPE & HES)

95% CI
The 95% confidence interval is the range around an estimate of some statistic (for example a mean, or a rate) such that there is a 95% (19 out of 20) chance that the interval contains the true value.

See
Randomised controlled trials in airways disease 2001/3
Ethnic variations in lower respiratory disease 2001/1

Age sex standardisation
The age-sex standardised rate for a particular condition is the rate which would have occurred if the observed age-sex specific rates for the condition had applied in a chosen standard population, for example in a standard European population. It is often applied to time trends in rates to take into account the changes in age and sex patterns over time in a population.

See
Trends in asthma mortality in Great Britain. 97/3

ICD
The International Classification of Diseases, published by the World Health Organisation, is a globally used classification system for cause of morbidity or mortality. Conditions are arranged by chapters. In the latest revision, ICD10, respiratory diseases are in chapter J. There are also diseases which affect the lung in other chapters such as respiratory tuberculosis (in chapter A, infectious and parasitic diseases), respiratory cancers (in Chapter C, Neoplasms) and cystic fibrosis (in chapter E, endocrine, nutritional and metabolic diseases).

See
Trends in emergency hospital admissions for lung disease. 2001/4
The burden of respiratory disease. 95/3

FEV1
This is the Forced Expiratory Volume in the first second. It is the volume of air that can be expelled in the first second of forced expiration (out-breath), starting from a full inspiration (in-breath). A low FEV1 is a marker of airway narrowing and is associated with various lung diseases.

See
Trends in COPD 2003/1

SMR
This can stand for Standardised Mortality Ratios or Standardised Morbidity Ratios, depending on the context. Other events can be standardised and given the term SERs. In all cases the principles are the same.

An SMR is often used to compare mortality (rate of deaths) in different sub-groups (for example populations in different regions, classes or ethnic groups). It takes into account the differing age-sex structures of the sub-populations being considered. A value of less than 100 shows lower than expected mortality and a value greater then 100 indicate higher than expected mortality.

The SMR is the ratio of the mortality actually observed in the sub-population to that expected if the sub-population had the same mortality risk as the standard population (such as that of England & Wales).

See
The geography of COPD mortality in the elderly. 96/1
Trends in asthma mortality in Great Britain. 97/3
Asthma and social class. 2000/3

Use of log scales
When plotting rates an absolute scale tends to accentuate changes in groups where rates are highest, for example, in plotting mortality where rates are highest in older age groups. A log scale can be used to enable trends in rates in different age groups to be compared. On a log scale the slope of the curve indicates the direction and magnitude of the change in rates.

See
Tuberculosis 2005/1
Trends in asthma mortality in the elderly 92/1

Hospital Admission data (HIPE and HES)
Until 1982, data on hospital admissions for England and Wales were available in the form of a random 10% sample of all discharges and deaths (Hospital In-Patient Enquiry - HIPE). These were collected under the Hospital Activity Analysis system (HAA). In 1982, Wales started collecting their data separately with a 100% sample. The 10% sampling continued in England until 1985.

In April 1987, changes were implemented in the NHS information systems in England following recommendations made by the Körner committee and the English data started to be collected under the Hospital Episode Statistics system (HES). Wales continued collecting data under the HAA system until 1991, when they too changed to the HES system in April 1991.

Under the HES system, discharges (and deaths) are identified as "finished consultant episodes" or FCEs. A consultant episode is a period of care spent under one consultant, and patients may experience more than one episode of care in an admission. Thus, the number of hospital episodes will over-estimate the number of hospital admissions.

In the early years of HES, the English data were published in books based on a 25% sample aggregated by financial year (April-March). English HES data from April 1989 onwards are currently available aggregated by financial year from the HES website. The data mostly refer to FCEs, though some limited data on admitting diagnosis are also available. The English HES data that LAIA present are based on a re-analysis of 100% HES data by calendar year. We also only use first consultant episodes, which represent the admitting diagnosis. English HES data were coded under WHO coding ICD9 until April 1995 when ICD10 was introduced.

The Welsh HES data continues to be collected separately from the English data and ICD10 was introduced for the Welsh HES data in April 1995. The Welsh HES data that LAIA presents is obtained from Health Solutions Wales. They re-aggregate the data for us by calendar year and is for first consultant episodes, which is the admitting diagnosis.

Scottish data on admissions are more complete. The numbers have been based on a 100% sample since 1968, and there has been no break or change in the data collection system. The data are collected as primary discharges from inpatient and day case episodes in general and acute wards and are now included as part of the Scottish Morbidity Record (SMR1). Data are collected in calendar years, and have been coded under ICD10 from 1997 onwards.

Hospital admission data in Northern Ireland are recorded on the Hospital Inpatients System.

See
Trends in Hospital Admissions and Deaths from Asthma. 2002/1
Trends in Emergency Hospital Admissions for lung disease. 2001/4
Trends in hospital admissions for asthma 96/2