Jacaranda Health

Our Recommendation

Jacaranda Health scores 100 out of 100, and it is the only organization we have reviewed that earned a perfect score while publishing randomized evidence that one of its two flagship programs largely did not work.

Jacaranda submitted both of its core programs to separate cluster randomized controlled trials across the same 40 Kenyan facilities. The PROMPTS digital platform improved maternal knowledge, birth preparedness, routine care seeking, and newborn care practices among 6,139 women, at 74 cents per woman enrolled. The MENTORS nurse mentorship program did not significantly improve provider knowledge or the quality of routine childbirth care, showed improvement in neonatal resuscitation that was not robust across sensitivity analyses, and showed no effect at all on severe maternal and neonatal morbidity and mortality: 10.3% in intervention facilities against 10.2% in control, with an adjusted effect of -0.008 whose confidence interval spans zero (P = 0.545).

That last measurement is why the score is 100. Jacaranda followed its own chain all the way to whether mothers and babies were harmed or died, measured it under randomization, and published the answer in a peer-reviewed journal when the answer was no.

Jacaranda Health's Fierce Certification score is 100/100 points based on our criteria:
✔ Has Ultimate Outcome Goals (20 pts)
✔ Measures Intermediate Outcomes (15 pts)
✔ Measures Ultimate Outcomes (20 pts)
✔ Measures Intermediate and/or Ultimate Counterfactual (20 pts)
✔ Shows Continual Learning & Adaptation (25 pts)

The Social Problem

Kenya's maternal mortality rate remains more than five times the United Nations Sustainable Development Goal, and high rates of severe maternal and neonatal morbidity persist across sub-Saharan Africa. What makes this problem distinctive is that it is no longer primarily a problem of getting women into facilities. Facility-based childbirth has risen substantially. Mortality has not fallen with it. The problem appears to have moved inside the building.

The failure has two sides. On the patient side, women often do not know the danger signs that should send them to a clinic, do not know when or where to seek care, arrive in labor with no margin left for complications, and skip postnatal care entirely. Only about half of Kenyan women receive any postnatal care.

On the provider side, the nurse who happens to be present when something goes wrong may not know what to do. Jacaranda's own trial found that providers performed only 58.6% of essential clinical actions during observed childbirths in its intervention facilities and 57.0% in control facilities. Routine maternity care quality was low everywhere.

Jacaranda's framing of what it is trying to produce is unusually direct for this sector: all mothers should experience a dignified childbirth, and all newborns should get a safe start in life.

The Solution

Jacaranda Health works through government public health systems, partnering with 24 county governments and the national Ministry of Health in Kenya. Its model has three components.

PROMPTS is a two-way SMS platform that enrolls pregnant and postpartum women at health facilities and sends stage-appropriate information, backed by a helpdesk that triages questions and refers women who need care. Jacaranda describes its purpose as empowering women to seek care at the right time and place and giving them greater agency in the health system. Over 1,200 facilities enroll women, with more than two million enrolled as of late 2023, at a cost of 74 cents per woman.

MENTORS trains frontline government nurses in basic and emergency obstetric and newborn care through continuing medical education lectures, simulation-based drills, and in-facility mentorship.

PULSE turns the data generated by PROMPTS and MENTORS into dashboards that county health managers and facility leaders can act on, showing where care is falling short in close to real time rather than at the end of a study.

The design logic is that patient-side demand and provider-side capability have to move together, which is also why Jacaranda evaluated them as one package and then analyzed them separately.

Key Outputs

• Mothers reached: More than two million women enrolled in PROMPTS as of November 2023, with over 1,200 facilities enrolling.
• Cost: 74 cents per woman enrolled in PROMPTS.
• Providers trained: Frontline government nurses trained in basic and emergency obstetric and newborn care across the partner network.
• Government partnership: 24 Kenyan county governments plus the national Ministry of Health, with expansion into Ghana, Tanzania, and Nigeria.
• Evaluation scale: Two cluster randomized controlled trials across 40 facilities in 8 counties, enrolling 6,139 women, interviewing 357 providers at baseline and 376 at endline, and directly observing 4,377 childbirths.
• Published strategy: A 2026 to 2028 Strategic Plan (PDF).
• New tools: DELTA, built to move from digital training to smarter support for frontline health workers.

Key Intermediate Outcomes

PROMPTS: knowledge and behavior (counterfactual). An unmasked, 1:1 parallel arm cluster randomized controlled trial in 40 facilities across 8 Kenyan counties, enrolling 6,139 pregnant women, with facility-level randomization and intention-to-treat analysis clustered by facility (PLOS Medicine). Outcomes were organized into six domains: knowledge, birth preparedness, routine care seeking, danger sign care seeking, newborn care, and postpartum care content.

Women enrolled in PROMPTS improved their knowledge and care-seeking behavior during both the antenatal and postnatal periods, alongside better newborn care practices (Innovations for Poverty Action).

The nulls are reported in the same paper. No statistically significant differences were found in the rate of facility-based childbirth, nor in the shares of women seeking medical advice or treatment for antenatal, postpartum, or neonatal concerns. No meaningful differences appeared in newborn physical exams or vaccines, where control-group rates were already 95% and 98%.

MENTORS: provider knowledge and care quality (counterfactual). An unblinded parallel arm cluster randomized trial in the same 40 facilities between November 2021 and August 2022 (PLOS Global Public Health).

Adjusted effects control for facility-level covariates and will not equal the difference between the two columns.

Jacaranda's own summary is that the program did not significantly improve provider knowledge or the quality of routine childbirth care, and that the neonatal resuscitation improvement was not robust across all sensitivity analyses.

Key Ultimate Outcomes

Severe maternal and neonatal morbidity and mortality (counterfactual, null).

Within the MENTORS cluster randomized trial, severe maternal and neonatal morbidity and mortality was 0.102 in control facilities and 0.103 in intervention facilities. The adjusted effect was -0.008, with a 95% confidence interval of -0.036 to 0.020 and P = 0.545. The raw means favor control by a thousandth; the adjusted effect favors the intervention by eight thousandths; the confidence interval contains zero. There is no detectable difference in either direction. Roughly one in ten births in both arms involved severe morbidity or death.

Two caveats. First, the trial ran from November 2021 to August 2022, which is a short window for a mortality outcome, and the study was designed primarily around quality of care. A null here is not proof that the program does nothing. Second, Jacaranda published it anyway, in a peer-reviewed journal, as a pre-specified outcome, alongside the null findings on provider knowledge and routine care quality.

The PROMPTS trial authors state the corresponding limitation in their own words: outcomes were self-reported and the study was not powered to detect effects on health outcomes, and future work is needed to ascertain the impact of at-scale digital platforms like PROMPTS on health outcomes.

What remains unmeasured. Dignity in childbirth and women's agency in the health system are the two ultimate outcomes Jacaranda names most prominently in its own mission language, and neither is measured.

Continual Learning & Adaptation

Jacaranda's learning behavior is the strongest we have reviewed, because it is the only organization that has published randomized evidence against itself.

• It randomized both of its flagship programs simultaneously, in the same facilities, and then analyzed them separately so the effects could be distinguished. The PROMPTS authors also named the risk that running both in the same facilities could have contaminated each other’s results, a threat no reader would have spotted on their own.
• It published the nulls prominently rather than burying them. The MENTORS paper's own abstract leads with the finding that the program did not significantly improve provider knowledge or routine childbirth care quality.
• It reported where its own effect was fragile. The neonatal resuscitation improvement was flagged as not robust across sensitivity analyses.
• It stated its own power limitation. The PROMPTS team wrote that the study was not powered to detect health outcomes and that future work is needed, which is an organization protecting readers from overreading its own result.
• It is testing the next iteration before scaling it. A further randomized controlled trial with Harvard, Rutgers, Jhpiego, and IPSOS is testing targeted postpartum contraception content, developed through human-centered design with early postpartum mothers, with the stated plan of integrating it into PROMPTS at full scale only if proven effective.
• It built new tools in response to what the trials showed. DELTA was built explicitly to move from digital training toward smarter support for frontline health workers, which is the problem the MENTORS trial identified.
• It built the data layer. PULSE exists so county health systems can see where care is failing rather than waiting for a trial.
• It published a forward strategy. The 2026 to 2028 Strategic Plan is public.

Further Reading

• Jacaranda Health homepage (jacarandahealth.org), Our Work (link), Impact and Insights (link)
• PROMPTS (link), MENTORS (link), PULSE (link)
• Vatsa, R., et al. “Impact evaluation of a digital health platform empowering Kenyan women across the pregnancy-postpartum care continuum: A cluster randomized controlled trial.” PLOS Medicine, 2025, 22(2): e1004527, February 3, 2025. DOI 10.1371/journal.pmed.1004527, PMID 39899612. The PROMPTS trial. Source of the 6,139 participants, the six outcome domains, the null findings on facility-based childbirth and care seeking, and the authors’ statement that the study was not powered to detect health outcomes. (link)
• Zhang, L., et al. “Impact of a long-term mentorship program for childbirth providers on quality of care: A cluster randomized controlled trial.” PLOS Global Public Health, 2026, 6(9): e0006073, September 2, 2026. DOI 10.1371/journal.pgph.0006073, PMID 42685042. The MENTORS trial. Source of the provider knowledge and care quality nulls, the neonatal resuscitation figures, the 57.0% and 58.6% essential actions rates, and the severe maternal and neonatal morbidity and mortality result. (link)
• Innovations for Poverty Action, The Impact of a Digital Health Platform on Maternal and Newborn Care in Kenya (link)
• Trial registration, Evaluation of a Nurse Mentorship and Digital Health Package Intervention in Kenya, NCT05110521 (link)
• Trial registration, Impact Evaluation of a Targeted mHealth Intervention to Improve Uptake of Postpartum Contraception in Kenya, NCT06266780 (link)
• Jacaranda Health 2026 to 2028 Strategic Plan (PDF)
• Why We Built DELTA (link)
• Strengthening Referral Systems to Save Mothers and Newborns Across Three Kenyan Counties (link)

• Share
• Tweet