Missed Opportunities for Stroke Prevention in Hypertensive Patients: A Retrospective Case-Control Study

This retrospective case-control study demonstrates that more frequent blood pressure monitoring and more intensive antihypertensive medication management are significantly associated with a reduced risk of ischemic stroke in hypertensive patients.

The Gist

High blood pressure is a major cause of ischemic stroke, a type of stroke where blood flow to the brain is blocked. While doctors know that lowering blood pressure can reduce this risk, many people with high blood pressure do not reach healthy target levels in daily clinical practice. This study investigated whether gaps in how patients are managed over time represent missed opportunities to prevent these strokes.

The researchers analyzed electronic health records from a large healthcare system involving over 40,000 adults with hypertension. To ensure a fair comparison, the study used a matched design. For every person who experienced an ischemic stroke, the researchers identified two similar people who did not have a stroke. These groups were matched by age, sex, race, and ethnicity. The researchers also aligned the timing of their observations so that both groups were monitored for the same amount of time.

The study looked at three specific ways that hypertension care is delivered: how often a patient has outpatient visits to measure blood pressure, how many different types of blood pressure medications they take, and how quickly a doctor adjusts treatment when blood pressure remains high.

The results showed that more active engagement in care was associated with lower odds of having a stroke. Specifically, patients who had four to five blood pressure measurements per year had lower odds of stroke compared to those who had only one visit per year. Regarding medication, patients taking one to three different types of blood pressure ingredients had lower odds of stroke than those taking none.

The most consistent finding involved how doctors responded to high blood pressure readings. The researchers used a score to measure "treatment intensification," which tracks how often a doctor starts a new medication, adds a new ingredient, or increases a dose in response to high blood pressure. The study found that higher intensification scores—meaning the treatment was adjusted more frequently and effectively in response to high readings—were associated with a steady decrease in stroke risk.

These patterns remained consistent even when the researchers looked at different groups of people, such as different ages, sexes, or races. The associations also held true regardless of whether a patient's blood pressure was moderately high or very high.

The researchers suggest that these findings indicate that gaps in routine management, such as infrequent monitoring or delayed changes to medication, may be missed opportunities for stroke prevention. The paper concludes that improving the consistency and responsiveness of hypertension care—through more regular follow-up and timely treatment adjustments—may be an important way to reduce the number of ischemic strokes.

Technical Summary

Technical Summary: Missed Opportunities for Stroke Prevention in Hypertensive Patients

Problem Statement

Hypertension is the primary modifiable risk factor for ischemic stroke, yet clinical management in routine practice is frequently suboptimal. While "therapeutic inertia"—the failure to escalate treatment despite uncontrolled blood pressure (BP)—is a documented phenomenon, most existing research focuses on intermediate clinical markers (like BP control levels) rather than definitive clinical endpoints (like ischemic stroke). Furthermore, previous studies have often failed to characterize hypertension management as a continuous, longitudinal care delivery process, making it difficult to determine if gaps in routine care (monitoring frequency, medication titration, and follow-up) directly translate to increased stroke risk.

Methodology

The researchers conducted a retrospective, 1:2 matched, nested case-control study using longitudinal Electronic Health Record (EHR) data from Sentara Health (a large regional health system in the US) spanning 2010–2024.

• Cohort & Matching: The study included 13,476 cases (patients with a first-ever incident ischemic stroke) and 26,952 matched controls. Matching was performed on age, sex, race/ethnicity, and calendar time.
• Temporal Alignment: To prevent bias from varying follow-up durations, the study used a "time-aligned" approach, where controls were assigned a pseudo-event date to ensure the observation window for both groups was identical. A 60-day "blanking period" was applied after the hypertension diagnosis to avoid reverse causation.
• Care Delivery Metrics: Three longitudinal dimensions of care were quantified:
  1. Monitoring Frequency: Number of outpatient visits with BP measurements per year.
  2. Medication Intensity: Number of antihypertensive medication ingredients used.
  3. Treatment Responsiveness: A "medication intensification score" (ranging from -1.0 to 1.0) that measures how effectively clinicians escalate therapy in response to episodes of uncontrolled BP (SBP $\ge$ 140 or DBP $\ge$ 90 mmHg).
• Statistical Analysis: Conditional logistic regression was used to estimate adjusted odds ratios (aORs), adjusting for baseline SBP/DBP, social vulnerability index (SVI), and various comorbidities (CHF, CKD, COPD, AF, diabetes, and dementia).

Key Contributions

1. Unified Care Construct: Unlike prior studies that looked at single metrics, this study characterizes hypertension management as a multi-dimensional, longitudinal care delivery pattern.
2. Hard Endpoint Validation: It directly links gaps in routine outpatient management to the occurrence of ischemic stroke, providing clinical evidence that management gaps are not just "suboptimal care" but "missed opportunities for prevention."
3. Methodological Rigor: The use of a nested case-control design with time-aligned observation windows minimizes the confounding effects of differential follow-up time.

Results

The study found that higher engagement in hypertension care was consistently associated with lower odds of ischemic stroke:

• Monitoring Frequency: Compared to $\le$1 visit per year, having 4–5 visits per year was associated with significantly lower stroke odds (aOR 0.43). Interestingly, the benefit attenuated for those with $>5$ visits, likely due to these patients having higher disease severity.
• Medication Ingredients: Using 2–3 medication ingredients was associated with reduced stroke odds (aOR 0.80) compared to no medication. However, using $>3$ ingredients did not show a significant additional reduction, potentially reflecting higher baseline disease complexity in those patients.
• Treatment Intensification: This was the strongest predictor. The highest quartile of the intensification score showed the most significant reduction in stroke risk (aOR 0.47).
• Subgroup Consistency: The protective associations were largely consistent across sex and SBP strata, though effect modification was noted in age (where high medication counts were associated with higher risk in patients $<45$ years) and SVI (where the benefits of frequent visits were weaker in high-vulnerability areas).

Significance

The findings suggest that the process of care is as critical as the level of BP control. Even if a patient's average BP is known, gaps in the frequency of monitoring and the speed of therapeutic response (intensification) represent actionable targets for stroke prevention.

Clinical Implications:

• Systemic Improvements: The study advocates for structured follow-up programs and standardized protocols for treatment escalation.
• Technology Integration: There is a clear opportunity to use EHR-based clinical decision support tools to prompt clinicians to intensify treatment when uncontrolled BP is detected.
• Health Equity: The variation in results across SVI strata highlights that social vulnerability may impede the effectiveness of routine care delivery, necessitating targeted interventions for high-risk populations.