Blood Pressure Digit Preference in Hypertensive Patients Attending Primary Health Care Is Associated With Future Cardiovascular Events

Objective: Digit preference refers to rounding blood pressure (BP) values more often than would occur by chance. Studies show that practices with high digit preference are linked to fewer antihypertensive prescriptions, and patients attending such practices have an increased cardiovascular (CV) event risk. The objective is to investigate how the fraction BP digit preference in hypertensive patients in primary health care centres relates to future CV events. Design and method: From the Swedish Primary Care Cardiovascular Database, 74,751 patients (41,728 women) in 48 primary care centers in Sweden with a recorded diagnosis of hypertension during 2001–2008 were included. The fraction of systolic and diastolic BP values ending in 5 or 10 during 2007–08 at each center was calculated. Cox regression analysis assessed the relation between digit preference at each primary health care center and CV events (ischemic heart disease, cerebrovascular disease, CV death) during 2009–2013, i.e. follow-up was 1–5 years. Covariates included age, gender, income, diabetes, ischemic heart disease cerebrovascular disease, and region. Results: Patient mean age in 2008 was 70±14 years. The fraction of digit preference varied between 32% and 100% between centres (Figure 1). There was a significant association between the utilization of digit preference at a primary health care centre and risk for future CV events for these patients (Figure 2). A proxy for quality of care by centre was estimated as fractions of patients with lipids, creatinine, and microalbuminuria assessed during 2007–2008. This revealed no association between markers of quality of care and prevalence of digit preference. Conclusions: BP digit preference is associated with future CV events. This may be attributed to inaccurate measurements procedures for BP measurements, faulty evaluation of actual BP levels and risk, treatment not offered appropriately, or other means of possible inappropriate quality of care.