What is the incidence of orthostatic hypotension in UK primary care? An electronic health record database study

Problem

Orthostatic hypotension (OH) is estimated to affect up to a third of community-dwelling older adults. OH and its resulting effect on reduced cerebral blood flow is linked to falls, fractures, ischaemic events, cognitive impairment and increased mortality. Older people with OH are 2.5 times more likely to have recurrent falls, compared to those without, costing the NHS > £2.3 billion. The potential benefits of routine screening for OH remains unclear and no studies have examined the incidence of symptomatic cases presenting to GP. This study aims to i) examine the incidence of OH over the last decade recorded in GP and ii) examine how trends in incidence of documented OH vary by age, sex and socio-demographic characteristics.

Approach

Longitudinal cohort study using the IQVIA medical research database – a primary care database of > 12 million patients. We included patients aged > 50 years, registered with a GP practice contributing data for at least 1 year between 2000 and 2018. Cases were identified as those with a new record of a Read code for OH. The recording of OH was estimated per 1,000 person years at risk (PYAR). A multivariable Poisson regression model was used to examine incidence by sex, age and social deprivation.

Findings

In total, 24,973 individuals (amongst 2,911,260 patients) had an electronic record indicating a new diagnosis of OH between 2008 and 2019. This was equivalent to 1.70 per 1,000 PYAR (95% CI 1.68-1.72). We found a higher incidence of OH with increasing age-band, with increasing social deprivation and in men compared to women. The incidence of OH increased at a greater rate by age-band amongst men, compared to women.

Consequences

The incidence of documented OH in GP between 2008-2018 was very low. There are no other studies examining incidence of OH in GP to make comparisons. However, a crude period prevalence of approximately 1% of coded cases amongst this dynamic cohort indicates this estimate is much lower than other studies which suggest community-dwelling prevalence to be closer to 20%. Low incidence in GP is likely due to a mixture of underreporting by patients where asymptomatic , under-detection and poor coding. Rising OH incidence with social deprivation is likely due to a greater prevalence of cardiovascular disease (CVD), multimorbidity and polypharmacy. Higher incidence of OH in men likely represents similarities in underlying pathology between OH and CVD. Further studies are needed to understand the significance of OH cases in GP and potential benefits of early identification.