(AN) Time Series Analysis of Blood Pressure-Lowering Drug Prescribing Trends Among Dementia Patients in Hong Kong

Publication Number: 656

Background: Ageing increases the risk of morbidities such as hypertension and dementia. Hypertension in the mid to late-life is associated with an increased risk of incident dementia and whether there are changes in prescribing patterns of blood pressure-lowering drugs with regards to a patient’s dementia diagnosis is unclear.

Objectives: The objective of the study is to evaluate whether there are changes in prescribing trends of blood pressure-lowering drugs before and after incident dementia diagnosis in Hong Kong.

Methods: This is a population-based study using electronic health data from Clinical Data Analysis and Reporting System (CDARS), a territory-wide medical database in Hong Kong. The target population were those aged 40 years or older diagnosed with dementia from 2004-2019. A time series analysis was conducted to compare the monthly proportion of patients prescribed with blood pressure-lowering drugs 3 years before and after incident dementia diagnosis using a segmented autoregressive error model. Eligible patients had a hypertension diagnosis and a blood pressure-lowering drug prescription before time series entry. Secondary outcomes were healthcare resource utilization (HRU) and uncontrolled blood pressure (UBP). Subgroup analysis was conducted by dividing patients into different comorbidity burdens based on Charlson Comorbidity Index (CCI): 0-3, 4-7 and 8+.

Results: The study included 31,873 patients with dementia in 2004-2019. The mean age at incident dementia diagnosis was 84.09 years old (Standard Deviation: 7.84). There was a decreasing prescribing trend of blood pressure-lowering drugs 3 years before dementia diagnosis, a sharp increase shortly after dementia diagnosis, then a greater decreasing trend in the 3 years after dementia compared to pre-dementia (Pre-dementia slope: -0.17, Level change: 4.66, Slope change: -0.30, all p-value < 0.01). For HRU, there was an increasing pre-dementia slope, a sharp increase shortly after dementia, then a gradual decreasing trend. For UBP, there was a decreasing pre-dementia slope, a decrease shortly after dementia, then a non-significant decreasing trend. The 0-3 and 4-7 CCI subgroups followed similar trends to the main analysis. The 8+ CCI group had a greater decreasing prescribing trend after dementia and higher overall HRU throughout the time series.

Conclusions: Patients with prior hypertension received increased medical care immediately after their first dementia diagnosis in Hong Kong. As their condition was better managed, the proportion of patients with uncontrolled blood pressure decreased, while prescribing and healthcare utilization in the 3 years after dementia diagnosis gradually decreased. Similar studies in other countries are warranted to support the current findings.