Clinical Outcomes Associated With a Remote Postpartum Hypertension Monitoring Program

Abstract

Objective

To evaluate differences in healthcare utilization and guideline adherence for postpartum individuals with a hypertensive disorder of pregnancy (HDP) engaged in a remote monitoring program compared with usual care.

Methods

This was a retrospective cohort study of postpartum individuals with HDP delivering from 3/2019–6/2023 at a single institution. The primary exposure was enrollment in a remote hypertension management program, that relies on patient home blood pressure (BP) measurement and centralized nursing team management. Patients enrolled in the program were compared with those receiving usual care. Outcomes included postpartum readmission, office visit within 6 weeks postpartum, BP measurement within 10 days, and initiation of antihypertensive medication. We performed multivariable logistic and conditional regression in a propensity score matched cohort. Propensity scores, generated by modeling likelihood of program participation, were assessed for even distribution by group, ensuring standardized bias <10% after matching.

Results

Overall 12,038 eligible individuals (6556 participants, 5482 controls) were included. Program participants were more likely to be White, commercially insured, be diagnosed with preeclampsia, and had higher prenatal and inpatient postpartum BPs. Differences in baseline factors were well-balanced after implementation of propensity score. Program enrollment was associated with lower 6-week postpartum readmission rates, demonstrating 1 fewer readmission for every 100 individuals in the program [PS-matched adjusted risk difference (aRD) −1.5 (−2.6, −0.46); adjusted risk ratio (aRR) 0.78 (0.65, 0.93)]. For every 100 individuals enrolled in the program, 85 more had a BP recorded within 10 days [PS-matched aRD 85.4 (84.3, 86.6] and 6 more had a 6-week postpartum office visit [PS-matched aRD 5.7 (3.9, 7.6)]. Program enrollment was also associated with increased initiation of an antihypertensive medication postpartum [PS-matched aRR 4.44 (3.88, 5.07)].

Conclusion

Participation in a postpartum remote BP monitoring program was associated with fewer postpartum hospital readmissions, higher attendance at postpartum visits, improved guideline adherence, and higher rates of antihypertensive use.

Précis:

Remote monitoring of postpartum hypertension was associated with fewer readmissions, increased adherence to national clinical guidelines, and initiation of antihypertensive medications for individuals with hypertensive disorders of pregnancy.

INTRODUCTION

Maternal morbidity and mortality are increasing in the United States with the majority of deaths occurring in the postpartum period.¹ Hypertension complicates 10–20% of pregnancies and is a significant contributor to postpartum morbidity and mortality.² Prior studies have shown worsening hypertension between days 3 and 6 postpartum, typically after discharge from the delivery hospitalization, which can be associated with serious morbidity, including seizure, stroke and rarely death.^3,4 Additionally, blood pressures can be labile during the postpartum period, necessitating frequent anti-hypertensive medication titrations.^5,6 As such, hypertensive disorders of pregnancy are the most common etiology of postpartum readmissions with a significant associated cost burden.⁷

The American College of Obstetricians and Gynecologists (ACOG) recommends that individuals with a hypertensive disorder of pregnancy (HDP) have their BP measured within 10 days after delivery.^2,9 However, prior studies have shown that attendance at these recommended in-person visits is low, with less than 50% of patients attending in-office BP checks.^10,11 More recently, with the growth of telemedicine, many institutions have developed and implemented remote BP management programs. A recent systematic review contracted by the Agency for Healthcare Research and Quality (AHRQ) demonstrated that home BP monitoring likely improves ascertainment of BP in the immediate postpartum period, however, the effect of these programs on postpartum outcomes is less clear.¹² The objective of this study was to evaluate outcomes following implementation of a remote monitoring program for postpartum individuals with a hypertensive disorder of pregnancy. Specifically, we sought to evaluate the association between enrollment in the remote monitoring program and postpartum care utilization, attendance at a postpartum office visit, and initiation of antihypertensive medications within 6 weeks postpartum.

METHODS

This was a retrospective cohort study including all birthing individuals eligible for our remote hypertension management program, delivering from March 2019 through June 2023 at Magee-Womens Hospital. Magee-Womens Hospital is the largest delivery hospital in Pennsylvania with an estimated 10,000 births each year, 30% of whom have a HDP and were eligible for remote blood pressure management. Eligibility for the program requires a diagnosis of a HDP during the delivery admission (preexisting or pregnancy-induced), or at least two inpatient postpartum blood pressures above 140/90 mmHg during the delivery hospitalization.¹³ Participants must speak English, Spanish or Portuguese and have a cellular device that accepts text messages. Individuals opt-in to the program and are enrolled either by manual entry by a physician or nurse or through recognition by an automated flag created in the electronic medical record to identify individuals with multiple elevated blood pressures during the delivery hospitalization. The automated flag is set up to send a daily list of individuals with elevated blood pressures to the nurse managers on the postpartum units to ensure all eligible individuals are offered participation in the program. Once enrolled, patients are provided with a blood pressure cuff and taught how to properly measure their blood pressure at home.

We designed nursing call center-driven blood pressure management and treatment algorithms that were developed by local expert stakeholders, consistent with national guidelines on goals for hypertension management postpartum.² Following discharge from the hospital, individuals are prompted via text to check their blood pressure at least daily for the first two weeks of the program and between 3–5 times per week for the remainder of the program through six weeks postpartum. Blood pressures are reported utilizing text messaging. Elevated BPs (≥140 /90 mmHg) or symptoms trigger automated alerts within the portal, which are reviewed by nursing staff and prompt physician contact based on our established.⁵ As there are no clear guidelines on BP thresholds for anti-hypertensive medication initiation or titration in the postpartum period, decision-making surrounding medication management is based on clinical judgment from the call center physician. Our program utilizes four call center physicians, all of whom are board-certified Maternal-Fetal Medicine physicians who have been overseeing the program since its inception in 2018. Individuals with symptoms including chest pain, severe headache, visual disturbances, shortness of breath or blood pressures ≥180 mmHg systolic or 120 mmHg diastolic are referred to the emergency department (ED) for further evaluation.

All data were extracted fromthe University of Pittsburgh Medical Center’s (UPMC) Clinical Data Warehouse (CDW) which stores all discrete documentation into the electronic health records. Baseline sociodemographic characteristics included year of delivery, race and ethnicity, maternal age, insurance, medical comorbidities, pre-pregnancy and time of delivery body mass index (BMI), area deprivation index for neighborhood of residence, parity, plurality and gestational age at first prenatal visit. Additional clinical characteristics included type of hypertension, receipt of magnesium sulfate, mode of delivery, and discharged on an anti-hypertensive medication. Inpatient and outpatient visit blood pressures were extracted from the CDW. Additionally, the CDW stored blood pressures that were patient-entered into a cell-phone interface through the third-party platform (Vivify). Blood pressures recorded in the hospital, at an office visit, or through the program were included in the analysis. Missingness of variables at prenatal visit are filled in with those populated at time of delivery when available. If they remain missing, these variables were replaced with the median value by exposure group. The remote monitoring program was approved by the UPMC Quality Improvement Review Committee and this research was reviewed by the University of Pittsburgh Institutional Review Board. Data collection and analysis were approved as an exempt study posing no greater than minimal risk; thus written informed consent was not required.

The primary exposure was enrollment in the remote monitoring program. As all individuals in this population are offered participation, we defined enrollment as accepting enrollment in the program and at least one remote BP entry postpartum. The primary outcome was postpartum care utilization, including hospital (readmission and ED visits) along with attendance of a postpartum office visit within 6 weeks postpartum. Secondarily, we assessed hypertension-specific outcomes, including adherence to ACOG guidelines of BP measurement within 10 days of delivery, and initiation of anti-hypertensive medications in the postpartum period. We compared eligible patients enrolled in the program wtih those who were not, using Pearson’s chi-square, two sample t-tests, and Wilcoxon rank-sum tests as appropriate. Because those who enrolled in the program were different at baseline from those who did not enroll (Table 1), we generated propensity scores using logistic regression to model likelihood of program enrollment. Propensity scores included demographic and clinical information obtained about the individual prior to program enrollment. Variables included in the propensity score were a combination of a priori decisions made by clinical experts, along with data-driven decisions from univariable analysis. Scores were evaluated for balance between groups.

Table 1.

Baseline sociodemographic and clinical characteristics of patients eligible for remote monitoring at Magee-Womens Hospital from 3/1/2019 through 06/30/2023 (n=12038), comparing those enrolled vs not enrolled in remote blood pressure monitoring program.

Factor	No Remote BP Monitoring n (%)	Remote BP Monitoring n (%)	Pvalueb
N	5482	6556
Delivery Characteristics
Delivery Year
2019	1468 (26.8%)	700 (10.7%)
2020	1187 (21.7%)	1493 (22.8%)	<0.001
2021	1149 (21.0%)	1646 (25.1%)
2022	1158 (21.1%)	1810 (27.6%)
2023	520 (9.5%)	907 (13.8%)
Self-reported race
Black	1357 (24.8%)	1333 (20.3%)	<0.001
White	3710 (67.7%)	4829 (73.7%)
None of the above	396 (7.2%)	362 (5.5%)
Unknown	19 (0.3%)	32 (0.5%)
Ethnicity
Not specified	529 (9.6%)	362 (5.5%)	<0.001
Hispanic or Latino	117 (2.1%)	111 (1.7%)
Not Hispanic or Latino	4836 (88.2%)	5961 (90.9%)
Maternal age (years)	30.3 (5.6)	30.7 (5.6)	0.002
BMI at time of Delivery (kg/m2) [Median (IQR)]	34 (29, 40)	35 (30, 40)	<0.001
Insurance
Commercial	2964 (54.1%)	4240 (64.7%)
Medicaid	2301 (42.0%)	2144 (32.7%)
Medicare	85 (1.6%)	79 (1.2%)
Self-Pay/ Other	132 (2.4%)	93 (1.4%)
Area Deprivation Index [Mean (SD)]	63.5 (28.5)	61.2 (28.3)	<0.001
Nulliparous	2720 (49.6%)	3805 (58.0%)	<0.001
PNC at UPMC	4964 (90.6%)	6135 (93.6%)	<0.001
Singleton Delivery	5270 (96.1%)	6236 (95.1%)	0.007
Gestational age at first visit (days) [Median (IQR)]	65 (56, 84)	64 (55, 78)	<0.001
Trimester at first visit
First	3967 (72.4%)	5186 (79.1%)	<0.001
Second	675 (12.3%)	625 (9.5%)
Third	307 (5.6%)	313 (4.8%)
Pregravid weight (lbs) [Median (IQR)]	170 (140, 210)	175 (145, 213)	<0.001
Morbidly obese	1232 (24.8%)	1610 (26.2%)	0.087
Chronic HTN	669 (13.5%)	791 (12.9%)	0.37
Any gestational HTN	2060 (41.5%)	3380 (55.1%)	<0.001
Preeclampsia	193 (3.9%)	587 (9.6%)	<0.001
Diabetes, Type 1 or Type 2	141 (2.8%)	203 (3.3%)	0.16
GDM	611 (12.3%)	778 (12.7%)	0.56
Anxiety	952 (19.2%)	1348 (22.0%)	<0.001
Depression	1063 (21.4%)	1416 (23.1%)	0.036
Cesarean delivery	2266 (41.3%)	2871 (43.8%)	0.007
Highest prenatal SBP (mmHg)	134 (128, 142)	139 (130, 146)	<0.001
Highest prenatal DBP (mmHg)	84 (80, 90)	88 (82, 94)	<0.001
Highest postpartum SBP within 2 hours of del (mmHg)	142 (133, 150)	147 (137, 158)	<0.001
Highest postpartum DBP within 2 hours of del (mmHg)	91 (85, 98)	93 (87, 101)	<0.001
Inpatient magnesium sulfate	511 (9.3%)	1898 (29.0%)	<0.001
SMM	375 (6.8%)	416 (6.3%)	0.27
Discharged on anti-hypertensive meds	459 (8.4%)	1840 (28.1%)	<0.001

^aMissingness of variables at prenatal visit are filled in with those populated at time of delivery when available.

^bPearson’s chi-squared, Two sample t test, Wilcoxon rank sum as appropriate

^cWithin 6 weeks of delivery.

BMI=body mass index; IQR=interquartile range; SD=standard deviation; PNC=prenatal care; HTN=hypertension; T1/T2=type 1 or type 2; GDM=gestational diabetes; SBP=systolic blood pressure; DBP=diastolic blood pressure; SMM=severe maternal morbidity; EDTR=emergency department treat and release; BP=blood pressure

We then conducted logistic regression modeling for each postpartum outcome, using three approaches to evaluate robustness of our results: (1) propensity score-matched conditional logistic regression, (2) propensity score-adjusted multivariable logistic regression, and (3) multivariable logistic regression adjusted for known confounders. Matching on likelihood of engagement was completed in a 1:1 ratio matching to nearest neighbor with a caliper of 0.036 without replacement. Results are reported as both risk-adjusted frequencies and associated difference for every 100 patients engaged in the program, along with risk ratios.

We performed a secondary analysis to evaluate blood pressure trajectories for participants in the remote monitoring program. Blood pressures at first prenatal visit, average inpatient BPs prior to delivery, average inpatient postpartum BPs, BPs closest to 7 days (within 2 days; i.e. between 5–9 days) and at 6 weeks (within 2 weeks; i.e. between 4–8 weeks postpartum) postpartum were used as the time points of interest. We limited this analysis to only those engaged in the program due to differential missingness of blood pressure recorded at 6 weeks when attempting to compare with those not enrolled (61% vs 80%, no remote monitoring vs remote monitoring; p<0.001), and because prior work has demonstrated that blood pressures recorded at office visits vary from self-measured home blood pressure values.¹⁴ This secondary analysis included only participants with data available at all 5 time points in care and was limited to BPs entered remotely.

All analyses were performed using Stata 16.0. A P value <0.05 was considered statistically significant.

RESULTS

In our cohort of 12,038 deliveries eligible for remote monitoring of hypertension, a total of 6566 (54.5%) enrolled in the program and had at least one remote blood pressure measurement. Rates of participation varied by year (Table 1). Individuals who enrolled in the program were more likely to be non-Hispanic White, commercially insured and reside in a more advantaged neighborhood. Clinically, individuals engaged in the program had higher prepregnancy weights and BMI at time of delivery, were more likely to be nulliparous, have prenatal care within UPMC that started earlier, be hypertensive in pregnancy (gestational hypertension 55.1% vs 41.5%, p<0.001; preeclampsia 9.6% vs 3.9%, p<0.11), deliver by cesarean, receive inpatient intravenous magnesium sulfate (29.0% vs 9.3%, p<0.001), and be discharged on an antihypertensive medication (28.1% vs 8.4%, <0.001). Both prenatal and inpatient systolic BP (SBP) (prenatal: 139 vs 134mmHg; p<0.001; inpatient: 147 vs 142mmHg; <0.001) and diastolic BP (DBP) (prenatal: 88 vs 84mmHg; p<0.001; inpatient: 93 vs 91mmHg; <0.001) were significantly higher in the enrolled group.

In univariable analyses, those enrolled in the remote monitoring program were more likely to attend an office visit within 6 weeks postpartum (77.0% vs 64.4%, p<0.001), be initiated on a new antihypertensive postpartum (24.4% vs 5.4%, p<0.001), have a BP recorded at 10 days or less from the time of delivery (97.6% vs 11.5%, p<0.001), and at 6 weeks postpartum (80.2% vs 60.9%, p<0.001). Five baseline variables had missing data that were equally distributed across groups: area deprivation index (5%), parity (6%), plurality (1%), weight (<1%) and height at time of deilivery (<1%).

Propensity scores were well balanced with 11,988 (99%) in the area of common support (Appendixes 1–3, available online at http://links.lww.com/xxx). Final propensity score models included delivery year month, maternal race and ethnicity, insurer, area deprivation index, BMI, parity, singleton vs multiple gestation, mode of delivery, induced delivery, highest inpatient postpartum BPs within 2 hours after delivery, having prenatal care within UPMC, pregnancy comorbidities (preexisting hypertension, preeclampsia, gestational hypertension, gestational diabetes, anxiety, depression, cardiomyopathy), delivering practitioner and severe maternal morbidity at time of delivery. When matched in a 1:1 ratio on propensity score, all variables had a standardized bias decreased to be <10% (Figure 1).

Figure 1.

Standardized bias for demographics and delivery characteristics before and after matching on propensity score. SBP, systolic blood pressure; DBP, diastolic blood pressure; HDP, hypertensive disorders of pregnancy; BP, blood pressure; BMI, body mass index; ADI, area deprivation index; GDM, gestation diabetes mellitus; HTN, hypertension.

In primary analyses, program participation was consistently associated with decreased hospital readmission. Results demonstrate 1 less readmission for every 100 patients engaged in the program [PS-matched adjusted RD −1.5 (−2.6, −0.46)], with no significant difference in ED visits (Table 2). There were also 6 more individuals attending an office visit within 6 weeks postpartum [PS-matched adjusted RD 5.7 (3.9, 7.6)]. Adherence to hypertension-specific guidelines increased indicating 85 more individuals with a blood pressure recorded within 10 days of delivery (PS-matched adjusted RD 85.4 (84.3, 86.6)) and 20 more initiated on an antihypertensive medication [PS-matched adjusted RD 20.0 (18.4, 21.5)] for every 100 engaged in the program (Table 2).

Table 2.

Risk-adjusted frequency of outcomes comparing those enrolled in a remote monitoring of postpartum hypertension program to those eligible but did not participate (N=12,038)

Model	No.	Total Visits/Total, No. (%)		Risk-Adjusteda frequency per 100 deliveries (95%CI)		Risk Difference, per 100 deliveries (95% CI)	Risk Ratio (95% CI)
		RM	No RM	RM	No RM
Readmission within 6 weeks
PS-matchedb	8166	219/4083 (5.4)	285/4083 (7.0)	5.4 (4.7, 6.1)	6.9 (6.2, 7.7)	−1.5 (−2.6, −0.46)	0.78 (0.65, 0.93)
PS-adjusted e	11988	371/6555 (5.7)	351/5433 (6.5)	5.4 (4.8, 5.9)	6.9 (6.2, 7.6)	−1.5 (−2.5, −0.61)	0.78 (0.67, 0.90)
Crude	12038	371/6556 (5.7)	354/5482 (6.5)	5.4 (4.9, 6.0)	6.8 (6.1, 7.5)	−1.3 (−2.3, −0.40)	0.80 (0.69, 0.93)
ED Visit within 6 weeks
PS-matchedb	8166	468/4083 (11.5)	459/4083 (11.2)	11.5 (10.5, 12.5)	11.2 (10.2, 12.2)	0.34 (−1.1, 1.7)	1.03 (0.91, 1.17)
PS-adjusted e	11988	758/6555 (11.6)	618/5433 (11.4)	11.6 (10.8, 12.3)	11.4 (10.5, 12.3)	0.15 (−1.1, 1.4)	1.01 (0.91, 1.13)
Crude	12038	758/6556 (11.6)	620/5482 (11.3)	11.7 (10.9, 12.5)	11.1 (10.3, 12.0)	0.60 (−0.62, 1.8)	1.05 (0.95, 1.17)
Office Visit within 6 weeks
PS-matchedb	8166	3014/4083 (73.8)	2722/4083 (66.7)	73.2 (71.9, 74.4)	67.4 (66.1, 68.7)	5.7 (3.9, 7.6)	1.08 (1.06, 1.11)
PS-adjusted e	11988	5044/6555 (77.0)	3509/5433 (64.6)	75.2 (74.1, 76.3)	67.1 (65.8, 68.3)	8.2 (6.4, 9.9)	1.12 (1.09, 1.15)
Crude	12038	5045/6556 (77.0)	3528/5428 (64.4)	74.8 (73.5, 75.5)	67.6 (66.5, 68.8)	6.9 (5.3, 8.5)	1.10 (1.08, 1.13)
BP Recorded within 10 days of delivery
PS-matchedb	8166	3990/4083 (97.7)	482/4083 (11.8)	97.6 (97.1, 98.1)	12.2 (11.2, 13.2)	85.4 (84.3, 86.6)	8.00 (7.35, 8.71)
PS-adjusted	11988	6400/6555 (97.6)	627/5433 (11.5)	97.6 (97.3, 98.0)	11.6 (10.7, 12.5)	86.1 (85.1, 87.1)	8.43 (7.79, 9.12)
Crude	12038	6401/6556 (97.6)	629/5482 (11.5)	97.5 (97.2, 97.9)	11.7 (10.8, 12.7)	85.8 (84.8, 86.8)	8.31 (7.67, 9.00)
Initiation of Antihypertensive Medications
PS-matchedb	8166	1023/4083 (25.1)	245/4083 (6.0)	25.8 (24.4, 27.1)	5.8 (5.1, 6.5)	20.0 (18.4, 21.5)	4.44 (3.88, 5.07)
PS-adjusted	11988	1599/6555 (24.4)	293/5433 (5.4)	24.5 (23.4, 25.6)	5.4 (4.8, 6.0)	19.1 (17.9, 20.4)	4.57 (4.04, 5.17)
Crude	12038	1599/6556 (24.4)	294/5482 (5.4)	24.9 (23.4, 26.0)	5.2 (4.6, 5.8)	19.7 (18.5, 21.0)	4.78 (4.23, 5.40)

^aAdjusted for delivery year month, maternal age, race, ethnicity, area deprivation index, parity, mode of delivery, commercial insurance, plurality, severe maternal morbidity, highest postpartum blood pressure at least 2 hours after delivery, administration of intravenous magnesium, body mass index at time of delivery admission, and prenatal care within UPMC.

^bMatched on propensity score including: delivery year month, race, ethnicity, area deprivation index, body mass index at time of delivery admission, parity, mode of delivery, induction, insurance type, SMM, plurality, highest and lowest postpartum blood pressure 2 hours after delivery, prenatal care within UPMC, prenatal diagnosis of diabetes, chronic hypertension, preexisting hypertension, preeclampsia, super imposed preeclampisa, gestational diabetes, anxiety, depression, cardiomyopathy, delivering provider and their specialty.

^cInclusive of blood pressures from 4–8 weeks with that closest to 6 reported.

Abbreviations: RM- remote monitoring, PS - propensity score, ED- Emergency Department, ADI- area deprivation index,

In the secondary blood pressure trajectory analyses, 4,334 patients enrolled in the program had at least 1 BP recorded prenatally, during delivery admission, at 7 days and 6 weeks postpartum. Trajectories demonstrate a marked increase in SBP at 1 week postpartum, that importantly returns to prenatal values by 6 weeks postpartum for those enrolled in the program (Figure 2; 121.1 mmHg vs 120.6 mmHg; paired t-test p=0.045). DBP trajectories show a similar spike at 7 days, though did not return entirely to prenatal measurements by 6 weeks postpartum (75.5 mmHg vs. 78.6 mmHg; paired t-test p<0.001).

Figure 2.

Blood pressure (BP) trajectory of initial prenatal blood pressure, average antenatal and postpartum blood pressure during delivery admission, postpartum remote blood pressure at one week, and postpartum remote blood pressure at six weeks for those engaged in a remote monitoring of postpartum blood pressure program (n=4,334). Mean systolic BP (mmHg) (A) and mean diastolic BP (mmHg) (B).

DISCUSSION

We found that participation in a postpartum remote BP management program was associated with fewer postpartum hospital readmissions, higher attendance at postpartum visits, improved national guideline adherence, and higher rates of antihypertensive use. Importantly, while we and others have previously shown that remote BP management improves ascertainment of blood pressure, this study demonstrates the downstream effects of such a program on clinical outcomes.^11–13,15 Additionally, blood pressure trajectories for those enrolled in the program demonstrated that SBPs returned to prenatal values despite a marked increase in the first 7 days postpartum. These findings have implications for reducing hypertension-related maternal morbidity and mortality in the postpartum period.

There are significant limitations to the postpartum care structures in the United States, namely that the majority of postpartum individuals do not attend visits after delivery.⁹ The major barriers to in-person care that exist in the postpartum period are only exacerbated among those in the most marginalized groups, who may have difficulties accessing transportation, obtaining child care or may have to return to work soon after delivery.^16,17 Blood pressure peaks on postpartum days 5–7, as previously reported and confirmed in our data, and these gaps in care in the postpartum period may lead to untreated severe hypertension and resultant hypertension-related morbidity.^3,18 Thus, innovative care in the postpartum period is essential to reducing maternal morbidity and mortality in the United States.

Home blood pressure monitoring has been consistently shown to improve adherence to ACOG guidelines, however, whether that translates to a reduction in maternal morbidity or improved outcomes has not been as well studied.^{10,12,13,15,19} In line with prior studies, we demonstrate that >95% of postpartum individuals participating in remote monitoring have a BP ascertained within the first 10 days after delivery. While improved ascertainment of blood pressure in this period is critical to identify hypertension, it is not necessarily sufficient to improve outcomes unless combined with access to systems for medication initiation and escalation of care if needed. We demonstrate that when compared to those who did not enroll, individuals who participated in our home management program were greater than 4-fold more likely to be initiated on anti-hypertensive agents in the postpartum period, which may translate to improvements in postpartum BP trajectory.

While those who participated had higher BP at the time of their delivery hospitalization, and clinical evidence of more severe hypertensive disease (i.e. more likely to be treated with magnesium, more likely to be discharged on anti-hypertensive medications), our program was associated with a return to early prenatal SBP by the six weeks postpartum. This improved BP control in the postpartum period has the potential to impact longer-term maternal cardiovascular health. The SNAP-HT trial and the POP-HT randomized postpartum individuals on anti-hypertensive agents after a hypertensive disorder of pregnancy to usual care with in-office blood pressure assessments versus home blood pressure monitoring plus management with systematic titration of anti-hypertensive medications in the postpartum period.^20,21 Both studies found improvements in diastolic blood pressure with a lowering of 4.5mmHg seen in the intervention group up to 6 months postpartum, improved cardiovascular remodeling and have more recently have demonstrated sustained effects up to 3–4 years postpartum.²²

The relationship between home BP management programs and care utilization is more complex. HDP are the most common reason for postpartum hospital readmission with significant healthcare and personal costs and implications on maternal-infant bonding.^7,23 However, increasing recognition of hypertension and contact with the healthcare system has the potential to lead to an increase in Emergency Department visits and postpartum hospital readmissions. While we demonstrate a reduction in postpartum hospital readmissions associated with our remote management program, our readmission rates are higher than previously published studies, which range from 1.5% to 5.0%.^10,11,15 This is likely related to baseline demographic differences and geographic variation in practice. In line with prior studies, we also demonstrate that program participants were more likely to attend 6-week in-person postpartum visits when compared to those who did not participate. While there may be significant upfront costs to establish remote monitoring programs, recent analyses have demonstrated that these programs are cost saving of $93 per patient and remain cost effective when the postpartum readmission rate is 3.0% or higher with standard monitoring.^24,25

Our study is strengthened by the large sample size and time period of 4.5 years. Our remote monitoring program incorporates standardized protocols for contact to the Maternal-Fetal Medicine physicians,⁵ minimizing variation—however, initiation of treatment is at the discretion of the on-call physician. Though our program currently enrolls >50% of the hypertensive population, future work should focus on optimizing engagement of all eligible individual, which s will likely include community-based interventions. We utilized several methods to address bias including propensity-score matching and propensity-score adjustment, though were not able to conduct a randomized controlled trial. We were also able to integrate data from the entire obstetric period, using prenatal and delivery admission data in conjunction with our remote monitoring blood pressures.

Our study has limited generalizability given that our data are from a single institution. Furthermore, we are limited to evaluating postpartum care utilization only within our health system. Additionally, we included all ED visits and rehospitalizations, not just those primarily related to hypertension in an effort to characterize healthcare utilization for other reasons that may indirectly be related to hypertension (such as heart failure or headaches). Finally, our program is only available to persons with access to a cellphone who speak English, Spanish or Portuguese. Lack of access to a cellphone, for roughly 10 individuals (annually) who opted out, may indicate higher social risks and these individual are unfortunately not captured in this program.

Our data adds to the growing body of literature supporting the critical role of remote BP management programs to improve adherence to guidelines and reduce postpartum hospital readmissions. Additionally, we show a favorable effect on postpartum BP trajectory among remote management participants. Remote management programs for postpartum hypertension show promise to bridge care gaps in the postpartum period, reduce disparities and potentially improve short and long-term maternal health.¹²