PREVALENCE OF DYSLIPIDEMIA IN NEWFOUNDLAND ADULTS: APPROACHES TO ESTIMATION USING ELECTRONIC MEDICAL RECORDS

Introduction Dyslipidemia is a leading risk factor for cardiovascular disease (CVD). Newfoundland and Labrador (NL) has a higher level of CVD mortality than any other province in Canada. This high level may be partially explained by the lipid profiles of people in this province. To our knowledge, there is no study in NL or Canada to use electronic medical records (EMR) to assess the prevalence of dyslipidemia. Objectives First, to assess the prevalence of dyslipidemia in NL using Canadian Primary Care Sentinel Surveillance Network (CPCSSN) EMR data. Second, to develop an algorithm that will provide a more accurate estimation of dyslipidemia using EMR data. Methods This is a secondary, cross-sectional analysis of existing data in our province. The study population included all patients aged 20 years or older who lived in NL. The most recent lipid profile (triglyceride, total cholesterol, high density lipoprotein (HDL-C), low density lipoprotein (LDL-C)) available on patients between 1 January 2009 and 31 December 2010 was identified. Independent variables included sex, age, lipid lowering medication use, and presence of comorbid conditions and other risk factors, such as hypertension and diabetes. The sensitivity, specificity, positive predictive value, negative predictive value and κ agreement, were calculated to compare different algorithms (ICD-9 code, laboratory result and lipid lowering medication use) for estimating the prevalence of dyslipidemia. Results This study included 4424 primary healthcare patients. Approximately 42% of patients were considered to have high total cholesterol, almost 36% had unhealthy levels of LDL-C, 25% had low HDL-C, and nearly 25% had high triglycerides. For adults with multiple dyslipidemias, elevated total cholesterol and LDL-C was the most common combination (32.9%), followed by elevated total cholesterol and triglycerides (13.1%). A combination of lipid lowering medication use and laboratory results revealed a sensitivity of 99.4%, and the κ agreement was 0.99 compared with the combination of the three existing variables. This algorithm showed a dyslipidemic prevalence of 76.0% among these patients. Conclusions Results of the NL component of the CPCSSN database showed a high prevalence of both individual and multiple dyslipidemias. Furthermore, the optimal criteria to estimate the prevalence of dyslipidemia in EMRs is using laboratory results together with lipid lowering medication use. These findings highlight the need for further investigation into lipid research in NL in order to determine the magnitude of dyslipidemia as an important risk factor of CVD and other chronic diseases in NL.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.