Hospital in-patient - Health-related outcomes
This category contains data on in-patient hospital admissions, obtained through linkage to external data providers. In-patients are defined as persons who are admitted to hospital and occupy a hospital bed. This includes both admissions where an overnight stay is planned and day cases.
"Spell and Episode Data" categories can be found in each of the five main sections of the in-patient data. A hospital Spell is the total time a patient is in hospital, from admission to discharge. An Episode is defined as a continuous period of care administered under one consultant within one hospital provider. If a patient is transferred to another consultant or to a different hospital provider during a Spell of treatment, a new record is generated, i.e. a new Episode.
In order to conduct analyses on the in-patient hospital data, you will need to select data-fields from the "Spell and Episode Data" sub-categories that interest you. For an overview of the amount and type of data stored across all Episodes for a particular data-field, please access the "Summary Information" sub-categories.
This data is large and complex and we advise you to read the explanatory documentation provided in the Additional Resources tab for details on how these data are presented and made accessible. The Essential Information
section presents summaries regarding the extent of data that is currently available, in the form of tabulations of the number of participants with one or more admissions for selected incident outcomes, and direct links to these are given below:
These numbers included in these tables are based on the first occurence of these health outcomes contained in hospital in-patient records since recruitment, and are not necessarily the first instance of the condition. The numbers of conditions occurring before recruitment will be made available in due course as additional health records are added to the core repository.
This data does not take into account information from participant self-report or other linked sources of data such as hospital outpatient records or primary care data. As such, outcomes not resulting in hospital admission are not captured by these data. Both because of this, and because UK Biobank participants are not representative of the general UK population, these numbers should not be regarded as estimates of disease rates in the UK population.
The conditions highlighted here are chosen based on their likely research interest and impact on public health and/or because they are the most common. In-patient data on cancer outcomes are not listed here because more complete data on cancer registrations is provided from linkage to cancer registry data (Category 100092).
The accuracy of the codes has not been checked in detail and UK Biobank's Outcomes Working group is currently working to adjudicate the most common outcomes.