Study Design
A retrospective cohort study was conducted among patients hospitalised with asthma to assess post-hospitalisation primary care management using routinely collected primary and secondary care data. We also assessed the association between patient characteristics and the likelihood of receiving post-hospitalisation asthma management in primary care. This was part of the Preventing Unscheduled Hospitalisations in Asthma (PUSH Asthma) study.11
Setting
We used primary care data from the Clinical Practice Research Datalink (CPRD) Aurum database, which covers 19% of the UK population and includes information on diagnoses (recorded using SNOMED-CT codes) and drug prescriptions. General practices in the UK which use the EMIS clinical information system contribute to this database. We also used linked Hospital Episode Statistics Admitted Patient Care (HES APC) data, which is coded using ICD-10 codes. Primary care data extraction was done using the Data Extraction for Epidemiological Research (DExtER) tool.12
Participants
Eligible patients were aged five years and older with a diagnosis of asthma prior to 1st January 2017 (index date), registered with a general practice contributing to CPRD Aurum at least one year prior to the index date, and hospitalised for asthma during the study follow-up period (1st January 2017 to 31st December 2019).
Asthma was defined as the presence of a SNOMED-CT code for asthma (see supplementary table 1 for SNOMED-CT codes). Asthma-related hospitalisation was defined by the presence of an ICD-10 asthma diagnosis code (J45- 46) as the primary diagnostic code in the linked HES APC data.
Patients who had diagnoses of Chronic Obstructive Pulmonary Disease (COPD), bronchiectasis, obstructive sleep apnea and Interstitial Lung Disease (ILD) were excluded from the analysis due to the potential for misclassification bias.
Baseline variables
Baseline data were extracted to describe demographic characteristics (age, sex, ethnic group, socioeconomic status measured by the Index of Multiple Deprivation [IMD] quintiles), behavioural risk factors (body mass index [BMI] and smoking status [for adolescents and adults only]) and relevant comorbidities (allergies, atopic eczema, allergic rhinitis, gastroesophageal reflux disease [GORD], chronic rhinosinusitis, anxiety, and depression) prior to the index date.
Asthma-related drug prescriptions within one year prior to the index date were extracted for the following medications: short-acting beta 2 agonists (SABA), oral corticosteroids (OCS), inhaled corticosteroids (ICS), long-acting beta 2 agonists (LABA), long-acting muscarinic antagonists (LAMA), leukotriene receptor antagonists (LTRA), and influenza vaccination.
Asthma-related hospitalisations within one year prior to the index date were also extracted.
Outcomes
The following aspects of asthma care were extracted from primary care records for within 48 hours, 7 days and 28 days of hospital discharge: provision of an asthma review, asthma management plan, prescriptions of asthma medications (SABA, OCS, ICS, LABA, LAMA, LTRA), demonstration of inhaler technique, and smoking cessation counselling. The primary outcome was a record of any of these items of care being recorded within 28 days following the date of hospital discharge.
We also assessed changes in inhaler medications post-hospital discharge by comparing the drug class of prescriptions recorded in the year before the hospital admission and those prescribed within 28 days after the hospital discharge.
Study size
We used the maximum number of eligible patients available in the database and study size was not determined by a formal sample size calculation.
Quantitative variables
Variables were categorised into the following groupings: age (5–11, 12–17, 18–24, 25–39, 40–59, 60–79 and ≥80 years), ethnic group (white, black, mixed, Asian, other, missing), IMD score quintile (1 [least deprived], 2, 3, 4, 5 [most deprived]), BMI (< 18.5kg/m2 [underweight], 18.5–24.9 kg/m2 [normal weight], 25-29.9 kg/m2 [overweight], > 30 kg/m2 [obese], missing), smoking status (current smoker, former smoker, never smoked, missing), SABA prescriptions (0, 1–3, 4–6, ≥7 prescriptions), and number of hospital admissions within the previous year (0, 1–3, 4–6, ≥7).
Statistical analysis
We stratified the cohort into children (5–11 years), adolescents (12–17 years) and adults (≥18 years). Baseline characteristics and outcomes for each age stratum were described using simple descriptive statistics. Missing data were addressed using a ‘missing’ category for each categorical variable. The primary outcome was presented as the proportion of hospitalised patients who received asthma management within 28 days of hospital discharge. Secondary outcomes were asthma management within 48 hours and 7 days of discharge. The associations between the primary outcome and patient characteristics were assessed using logistic regression, adjusted for demographic and clinical factors. Data analyses were done using Stata SE V.16 and R Studio.