PINGS
Phone-Based Intervention Under Nurse Guidance for Control of Hypertension After Stroke: A Randomized Multicenter Phase 3 Trial in Ghana
Bottom Line
A pragmatic, low-tech, nurse-navigated, phone-based mHealth bundle improved 12-month systolic BP control after stroke in Ghana (67% vs 43% achieving SBP <140 mm Hg, between-group difference 24%, p<0.001), without translating into significant reductions in major adverse cardiovascular events or mortality at one year.
Major Points
- Phase III, multicenter, open-label, blinded-endpoint RCT across 10 hospitals in Ghana (3 tertiary, 2 secondary/district, 5 primary), enrolling 500 recent stroke survivors with uncontrolled BP โฅ140/90 mm Hg.
- Intervention combined weekly home BP self-monitoring with nurse-navigator follow-up for out-of-range readings, daily phone-alarm medication reminders, and once-weekly 1-2 minute audio health-education messages in local Ghanaian dialects for 48 weeks.
- Primary outcome (proportion with SBP <140 mm Hg at month 12) markedly favored PINGS: 67% vs 43% (ฮ 24%, 95% CI 15-33%, p<0.001 by ITT; 71% vs 45% per-protocol, p<0.001).
- Mean systolic BP fall from baseline at 12 months was 20 vs 14 mm Hg (between-group difference -5.9 mm Hg, 95% CI -11.0 to -1.1, p=0.016).
- BP <140/90 mm Hg at 12 months: 56% vs 38% by ITT (60% vs 41% per-protocol), p<0.001.
- No significant differences in MACE (adjusted HR 0.79, 95% CI 0.35-1.79), all-cause mortality (5.7% vs 5.5%), serious adverse events (11.1% vs 7.0%, p=0.12), or modified Rankin Scale shift at 12 months (OR 1.12, 95% CI 0.79-1.57).
- Mediator analyses showed no detectable difference in medication adherence (medication possession ratio, Hill-Bone), hypertension self-care, health literacy, or quality of life; total antihypertensive treatment intensity score was modestly higher in PINGS over follow-up.
- Effect of intervention preserved across subgroups including stroke type, sex, age, education, income (effect actually larger in lower-income participants), baseline SBP, and across primary/district vs tertiary hospital sites โ supporting scalability in resource-limited settings.
Design
Study Type: Randomized Controlled Trial
Randomization: 1
Blinding: Open-label with blinded outcome assessment (PROBE design); office BP measured by trained assessors blinded to allocation; independent panel adjudicated MACE.
Enrollment Period: October 23, 2020 - April 21, 2023
Follow-up Duration: 12 months (last follow-up 5 April 2024)
Centers: 10
Countries: Ghana
Sample Size: 500
Analysis: Intention-to-treat (primary); per-protocol sensitivity; missing primary-outcome data imputed by last observation carried forward; analyses adjusted for baseline imbalances in stroke type and diabetes.
Inclusion Criteria
- Age โฅ18 years
- Head CT-confirmed ischemic or intracerebral hemorrhagic stroke within 1 month of onset (or positive 8-item Questionnaire for Verifying Stroke-Free Status [8-QVSFS] when neuroimaging unavailable)
- Uncontrolled blood pressure โฅ140/90 mm Hg at screening (mean of 3 readings)
- Owns a basic mobile phone or smartphone capable of receiving text/audio messages
- Able to provide written informed consent (self or reliable proxy)
Exclusion Criteria
- Any condition limiting cooperation with home BP monitoring/follow-up (e.g. severe cognitive impairment)
- Pregnant or nursing women
- Prisoners
- Institutionalized individuals
Arms
| Field | PINGS Intervention | Control |
|---|---|---|
| Intervention | Usual care plus PINGS bundle: (1) home automated BP monitor (Omron version 10) with โฅweekly self-measurement and nurse-navigator phone call if home BP >180/105 or <100/75 mm Hg; (2) daily medication-reminder phone alarm set by nurse navigator; (3) weekly 1-2 min audio health-education messages in preferred Ghanaian dialect delivered via Viamo for 48 weeks. Nurse navigators trained for 80 hours and used protocolized algorithms; uncontrolled home readings triggered clinic visit for treatment intensification under physician supervision. | Standard local post-stroke secondary prevention per Ghanaian guidelines with clinic visits every 2 months. Attention-control SMS texts on healthy lifestyle (not on medications, hypertension, or stroke). |
| Duration | 12 months | 12 months |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Proportion of participants achieving systolic BP <140 mm Hg at month 12 (intention-to-treat) | Primary | 43% (109/256) | 67% (163/244) | <0.001 | |
| Mean change in SBP from baseline at 12 months (mm Hg) | Secondary | -14 (SD 26) | -20 (SD 24) | 0.016 | |
| BP <140/90 mm Hg at 12 months (ITT) | Secondary | 38% | 56% | <0.001 | |
| BP <140/90 mm Hg at 12 months (per-protocol) | Secondary | 41% | 60% | <0.001 | |
| Major adverse cardiovascular events | Secondary | Reference | PINGS | 0.79 | NS |
| All-cause death | Secondary | 5.5% (14/256) | 5.7% (14/244) | NS | |
| Modified Rankin Score shift (baseline to 12 mo) | Secondary | Reference | PINGS | 1.12 | 0.52 |
| Hypertension self-care profile | Secondary | Comparable | Comparable | NS | |
| Medication possession ratio | Secondary | Comparable | Comparable | NS | |
| Hill-Bone hypertension compliance | Secondary | Comparable | Comparable | NS | |
| Hypertension/stroke health literacy | Secondary | Comparable | Comparable | NS | |
| Health-related quality of life (EQ-5D) | Secondary | Comparable | Comparable | NS | |
| Mean antihypertensive therapeutic intensity score over 12 mo | Secondary | 1.38 (SD 0.74) | 1.49 (SD 0.76) | <0.01 | |
| Serious adverse events | Adverse | PINGS 11.1% (27/244) vs UC 7.0% (18/256), p=0.12 | |||
| Hospitalizations | Adverse | PINGS 2.9% (7/244) vs UC 1.2% (3/256), p=0.21 | |||
| Deaths | Adverse | PINGS 5.7% (14/244) vs UC 5.5% (14/256) | |||
Subgroup Analysis
Treatment effect on primary outcome consistent across age (<60 vs โฅ60), sex, education, baseline SBP (<160 vs โฅ160 mm Hg), diabetes status, stroke type (ischemic vs hemorrhagic), eGFR strata, and across hospital level (primary/district RR 1.50, 95% CI 1.19-1.87; tertiary RR 1.66, 95% CI 1.29-2.13). Effect was numerically larger in lower-income participants (<500 GHc/month) than higher-income, suggesting the intervention may narrow income-based hypertension control gaps.
Criticisms
- Open-label design โ patients, clinicians, and on-site research staff were unblinded; only outcome assessors and MACE adjudicators were blinded (PROBE design).
- 12.3% of PINGS and 12.5% of UC participants did not complete the study; missing primary outcomes imputed by last observation carried forward, which can bias toward earlier (less controlled) measurements.
- Clinic-measured (not ambulatory) BP at screening may have excluded patients with masked hypertension and let in some with white-coat BP elevation.
- Sample size (n=500) and 12-month follow-up underpowered to detect differences in recurrent stroke, MACE, or mortality.
- Stroke-type and diabetes imbalance at randomization required statistical adjustment.
- Pre-specified mediators (medication adherence, self-care, health literacy, quality of life) did not differ, leaving the mechanism of BP improvement only partially explained; total antihypertensive intensity score was modestly higher in PINGS at baseline as well as during follow-up.
- Generalizability beyond Ghana to other low- and middle-income settings unproven; future hybrid implementation studies with longer follow-up needed.
Funding
Trial conducted under Ghana-US collaboration with ethical oversight from Kwame Nkrumah University of Science and Technology (Kumasi, Ghana) and the Northern California Institute of Research and Education (NCIRE), USA. (Full funding sources per published paper.)
Based on: PINGS (Circulation, 2026)
Authors: Sarfo FS et al.
Citation: Circulation. 2026 Jun 16. doi:10.1161/CIRCULATIONAHA.125.077424
Content summarized and formatted by NeuroTrials.ai.