Hypertensive disorders of pregnancy (HDP) affect up to 10% of pregnancies and encompass a spectrum of disorders that includes preexisting chronic hypertension; gestational hypertension; preeclampsia with and without severe features; eclampsia (seizures); and hemolysis, elevated liver enzymes, and low platelet count (HELLP) syndrome.¹ Historically, it was believed that HDP was cured by delivery of the placenta, but it is now understood that HDP can persist, worsen, or develop de novo after birth, and may result in severe morbidity or mortality due to eclampsia and stroke.^2,3 Beyond the postpartum period, data suggest that individuals with pregnancies complicated by HDP have a higher risk of chronic hypertension and a higher lifelong risk of cardiovascular complications.^4–6

Hypertensive disorders of pregnancy and their sequelae disproportionately affect minoritized people from some racial and ethnic groups and communities marginalized by geography, race, and socioeconomic status.^7,8 The prevalence of HDP is highest in people of non-Hispanic Black, American Indian, and Alaska Native backgrounds.⁹ Overall, Black people are three times more likely than non-Hispanic White people to die of pregnancy-related conditions, both around the time of delivery and up to 1 year postpartum.⁷ A higher percentage of these deaths are attributable to people with HDP who are non-Hispanic Black (9.9%) compared with non-Hispanic White (4.8%).¹⁰

Some individuals with HDP are discharged from postpartum hospitalization with antihypertensive medications. Others may develop hypertension after discharge and require treatment. In the postpartum period, blood pressure (BP) can change rapidly and unpredictably (due to labile postpartum physiology) before returning to normal, and this can result in shifting medication requirements. Home BP monitoring may allow earlier detection of postpartum hypertension or preeclampsia² while also improving the patient experience by increasing the convenience of care and decreasing the frequency of clinical encounters.

This systematic review sought evidence regarding 1) the effectiveness and harms of home BP monitoring, with or without remote monitoring, compared with traditional clinic-based BP checks; and 2) the comparative effectiveness and harms of various home BP–monitoring implementations (eg, monitoring regimen, communication method) in improving clinical outcomes and reducing existing disparities.

This report is derived from a more comprehensive systematic review of management of HDP that was nominated by the American College of Obstetricians and Gynecologists, funded by the Patient-Centered Outcomes Research Institute, and performed by the Brown Evidence-based Practice Center under contract to the Agency for Healthcare Research and Quality (AHRQ).¹¹ The complete review was published on the AHRQ website concurrently with this publication (https://effectivehealthcare.ahrq.gov/products/hypertensive-disorders-pregnancy/research)

SOURCES

We used methods consistent with those outlined in the Methods Guide for Effectiveness and Comparative Effectiveness Reviews (available athttps://effectivehealthcare.ahrq.gov/topics/cer-methods-guide/overview), in accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses).¹² The PROSPERO protocol registration number is CRD42022313075.

Literature searches (for the larger AHRQ review) were conducted in Medline (via PubMed), the Cochrane Register of Clinical Trials, the Cochrane Database of Systematic Reviews, EMBASE, CINAHL, andClinicalTrials.gov, from inception to December 1, 2022. The reference lists of relevant existing systematic reviews were screened for additional eligible studies. See Appendix 1, available online athttps://links.lww.com/AOG/D229, for detailed search terms.

STUDY SELECTION

Based on a priori criteria, we included randomized controlled trials (RCT)s and nonrandomized comparative studies that compared home BP monitoring during the postpartum period (up to 1 year after delivery) with clinic-based BP monitoring, other nonclinic approaches (eg, kiosks, pharmacy-based), or alternative home BP–monitoring interventions (including alternative training, education, or alert-triggering protocols). To provide additional context regarding intervention components and implementation strategies, we included single-arm studies of 50 or more participants who were offered home BP monitoring. With input from key informants and technical experts (named in the full AHRQ report), we selected prioritized outcomes related to BP management, severe maternal health outcomes, patient-reported outcome measures,¹³ patient-reported experience measures,¹⁴ health care utilization, and harms. We also sought analyses that evaluated health care disparities based on race and ethnicity, socioeconomic status, insurance coverage, and other social determinants of health.

Abstracts were screened in duplicate using Abstrackr software, which incorporates machine learning.¹⁵ Conflicts were resolved by consensus of the systematic review team. Potentially relevant citations were retrieved in full text and rescreened in duplicate.

Data on study design, baseline characteristics, interventions, outcomes, and results were extracted into SRDR+ (https://srdrplus.ahrq.gov/). Each eligible study was extracted and assessed for risk of bias or quality by one researcher and confirmed by a second, independent researcher. For RCTs, we used items from the Cochrane Risk of Bias tool.¹⁶ For nonrandomized comparative studies, we used relevant items from the Cochrane tool and specific sections of the ROBINS-I tool¹⁷ that pertain to confounding and selection bias. For single-arm studies, we assessed relevant methodologic quality items from the Cochrane tool and the NHLBI Tool¹⁸ but did not assign an overall risk of bias.

We calculated the between-intervention effect sizes using the following effect measures: relative risk (RR) for common dichotomous outcomes¹⁹ and mean difference (between-group difference) for continuous outcomes. When a study had no events (outcomes) in one group, we calculated the risk difference. Adjusted analyses were preferentially extracted over unadjusted (crude) data. Meta-analysis was performed if at least three studies reported sufficiently similar outcomes.

We graded the strength of the body of evidence provided by RCTs and nonrandomized comparative studies, per AHRQ methods,²⁰ by considering the number of studies and study participants, the study limitations (accounting for both study design and risk of bias), the directness of the evidence to the question at hand, the consistency of study results, the precision of any estimates of effect, and other limitations (particularly sparseness of evidence, including rare events). Strength-of-evidence ratings were assigned as high, moderate, low, or insufficient to estimate an effect. In accordance with AHRQ guidance, we use qualifying language regarding strength of evidence when communicating conclusions: “probably” for moderate strength and “may” for low strength.^21,22

RESULTS

Searches for the full review yielded 16,105 records, of which 380 were retrieved and screened in full text. Thirteen studies (in 18 articles^23–40) pertained to home BP monitoring: three RCTs, two nonrandomized comparative studies, and eight single-arm studies (seeFig. 1 for the literature flow diagram and Appendix 2, available online athttps://links.lww.com/AOG/D229, for design details). The studies were published between 2017 and 2022 and included 3,867 participants.^23–40 Eleven studies were conducted in the United States, one was conducted in the United Kingdom, and one was conducted in India. The three RCTs had a total of 510 participants, the two nonrandomized comparative studies had 901 participants, and the eight single-arm studies had 2,456 participants. All of the comparative studies enrolled participants with a prior HDP diagnosis. The proportion of participants who were Black among the U.S. comparative studies ranged from 28.6% to 76.3%. The evidence was insufficient to conduct meta-analyses. No studies evaluated nonclinic approaches (eg, kiosks, pharmacy-based care). A single study with partial results reported inClinicalTrials.gov evaluated the comparative effectiveness of remote monitoring and discharge with a BP cuff.³⁹

We assessed the overall risk of bias as moderate for the three RCTs (due to lack of blinding of participants and outcome assessor) and for one nonrandomized comparative study³⁷ (possible residual confounding) and high for the other nonrandomized comparative studies³² (due to potential for confounding resulting from inclusion of participants in the standard care arm who had declined to participate in the home BP–monitoring intervention).

Overall, there is moderate strength of evidence that home BP monitoring improves reporting of BP measurements (seeFig. 2 for summary andTable 1 for detailed evidence profile). Two studies found a reduced likelihood of hypertension-related hospital admissions with home BP monitoring compared with usual care (low strength of evidence). Black participants in the home BP–monitoring arms of two studies reported BPs at rates similar to non-Black participants, whereas Black participants assigned to office-based follow-up were less likely than non-Black participants to attend scheduled follow-up appointments that included BP checks. Based on these findings, we conclude that home BP monitoring probably reduces racial disparities in BP ascertainment (moderate strength of evidence).

Three studies (two RCTs and one nonrandomized comparative study) compared home BP monitoring with usual care (Table 2). Cairns et al^24,25 randomized 91 participants to home BP monitoring with supervised self-management of antihypertensive medication compared with usual care. The experimental group transmitted home BP readings by smartphone application or text message. Patients were automatically sent instructions about medication titration that incorporated an individualized medication-reduction schedule. Those allocated to usual care had their BP monitored by their community midwife and their antihypertensive medication adjusted by their general practitioner. Median (interquartile range) treatment duration (the time until patients were able to cease antihypertensive medication treatment) was 29 days (12–49 days) in the intervention group and 41 days (23–58 days) in the control group (adjusted mean difference −12 days, 95% CI −39 to 6). Participants who self-managed had significantly lower mean diastolic BP at 4 weeks (adjusted mean difference −3.0 mm Hg, 95% CI −5.8 to −0.1), 6 weeks (adjusted mean difference −5.8 mm Hg, 95% CI −9.1 to −2.5), 12 weeks (−4.3 mm Hg, 95% CI −7.7 to −0.5), and 26 weeks (adjusted mean difference −4.5 mm Hg, 95% CI −8.1 to −0.8). The reductions in diastolic BP persisted despite progressively fewer participants remaining on antihypertensive medications.

A follow-up on the same study, Kitt et al,²⁶ included 63 (69.2%) participants from the original trial who were evaluated with 24-hour BP measurements. Participants who had self-managed their BP continued to have significantly lower diastolic BP at 3- to 4-year follow-up than those in the usual-care group (adjusted mean difference −7.4 mm Hg, 95% CI −10.7 to −4.2). Participants in the experimental group were significantly more likely to report higher (better) scores (on a 5-point Likert scale) when asked, “How much in control do you feel of managing your condition?” The mean score was 4.8 in the intervention group compared with 3.9 in usual-care group (adjusted mean difference 0.7, 95% CI 0.3–1.2).

Hirshberg et al³⁰ randomized 206 individuals with pregnancy-related hypertension to home BP monitoring by bidirectional text messaging or to office-based follow-up 4–6 days postpartum. Participants in the intervention group were automatically reminded to text their BP measurements twice daily for 2 weeks postpartum. During the first 10 days postpartum, significantly more participants in the texting group than in the usual-care group had a least one BP measurement (95/103 [92.2%] vs 45/103 [43.7%]); after adjusting for age, race, insurance, body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), parity, disease severity, mode of delivery, and chronic hypertension, the adjusted odds ratio was higher for the texting group (58.2, 95% CI 6.2–208.1). Outpatient antihypertensive medications were initiated using a standardized algorithm in 17 of 103 (16.5%) individuals in the intervention group and 10 of 45 (22.2%) individuals in the office-visit group who reported a BP measurement. Missing data for those not attending an office visit resulted in an imprecise, and potentially biased, effect estimate (adjusted odds ratio 1.0, 95% CI 0.3–3.1). There were no hypertension-related readmissions (0/103) in the home BP monitoring group and 4 of 103 (3.9%) in the usual-care group (risk difference −3.9%, 95% CI −7.6 to −0.02). Patients assigned to home BP monitoring rated the importance of face-to-face communication significantly lower than those in the office-visit group (P=.02). Overall, 141 of 206 (68.4%) study participants were Black. In the usual-care arm, 21 of 30 (70.0%) non-Black participants returned for a BP visit compared with 24 of 73 (32.9%) Black participants (RR 2.13; 95% CI 1.42–3.19). In the home BP–monitoring arm, 32 of 35 (91.4%) non-Black participants reported at least one BP, compared with 63 of 68 (92.6%) Black participants (RR 0.99, 95% CI 0.87–1.11).³¹ The ratio of relative risks⁴¹ in the two arms (ie, 0.99 in the home BP–monitoring arm divided by 2.13 in the office-based follow-up arm) was 0.46 (95% CI 0.30–0.71), which can be interpreted as an approximately 50% reduction in racial disparity due to the home BP–monitoring intervention.

Hoppe et al³² performed a single-center, prospective, nonrandomized comparative study in which they allocated 428 patients to remote BP monitoring or standard care, with clinic visits at 7–10 days and 6 weeks postpartum. Patients were matched by similar HDP diagnosis, but the study was at high risk of bias due to potential for confounding resulting from inclusion of participants in the standard-care arm who had declined to participate in the home BP–monitoring intervention. Participants in the home BP–monitoring arm were asked to submit BP readings for 6 weeks using a Bluetooth-enabled BP monitor. Significantly more participants in the telehealth arm had BP measurements within 10 days of delivery (202/214 [94.4%] vs 129/214 [60.3%] in the usual-care arm; adjusted relative risk [aRR] 1.59, 95% CI 1.36–1.77). In the unadjusted analysis, significantly more participants in the telehealth arm (57/214 [26.6%] vs 37/214 [17.3%]) received antihypertensive medications based on a standardized algorithm (RR 1.54, 95% CI 1.06–2.23). However, after adjusting for delivery mode, insurance status, antihypertensive medication use at time of discharge, and duration of postpartum admission, this effect was no longer statistically significant (aRR 1.03, 95% CI 0.74–1.44). Participants in the telehealth arm had significantly fewer hypertension-related hospital readmissions (aRR 0.12, 95% CI 0.01–0.96).

Two studies (one RCT and one retrospective pre–post nonrandomized comparative study) compared alternative home BP–monitoring strategies (Table 2). In both studies, all participants were discharged with a BP cuff. In the usual-care arms, participants were asked to keep a paper log. In the intervention arms, participants’ BPs were transmitted to an internet-connected tablet or reported during a scheduled audio-only telehealth visit. A single-center RCT³⁹ with partial results reported inClinicalTrials.gov randomized 213 participants to semi-automated or usual-care paper recording of home BP–monitoring measurements. Participants in the semi-automated monitoring group were given a Bluetooth-enabled BP cuff that transmitted BP measures from the monitor to a tablet. Those in the usual-care group received a prescription for a BP cuff and were asked to keep a paper log. Participants were asked to monitor their BP twice a day for 14 days after discharge. Adherence to BP reporting (the percentage of the 28 requested BP measurements reported) was substantially higher in the intervention group (with a Bluetooth-enabled BP cuff), with a median of 61.1% adherence (range 0.0–92.3%), compared with a median of 0.0% (range 0.0–100%) in the usual-care group. In other words, more than half of the participants in the paper-log group did not report any home BP measurements. Among patients discharged without antihypertensive therapy, those in the intervention group were significantly more likely to have hypertension diagnosed, with subsequent prescription of antihypertensive medication (24/60 [40.0%] in the intervention vs 10/65 [15.4%] in the usual-care group; RR 2.60, 95% CI 1.36–4.98). In the intervention group, 38 of 101 patients (37.6%) were referred to a primary care physician for continued BP management, compared with 31 of 112 (27.7%) in the control group (RR 1.36, 95% CI 0.92–2.01). Among patients in the intervention group, 25 of 101 (24.8%) had an emergency department visit compared with 17 of 112 (15.2%) in the control group (RR 1.63, 95% CI 0.94–2.84). Among patients in the intervention group, 13 of 101 (12.9%) were readmitted compared with 7 of 112 (6.3%) in the control group (RR 2.06, 95% 0.86–4.96). A larger percentage of preeclampsia-related complications was detected in the intervention group (9/101 [8.9%] vs 4/112 [3.6%]; RR 2.50, 95% CI 0.79–7.85). There were no maternal deaths in either group.

A single-center, retrospective, nonrandomized comparative study³⁷ evaluated 473 patients with HDP discharged during two time periods: pretelehealth and postinitiation of audio-only telehealth during the coronavirus disease 2019 (COVID-19) pandemic. All patients were managed according to the STAMPP HTN (Systematic Treatment and Management of PostPartum Hypertension) quality-improvement bundle, given home BP cuffs at discharge, and instructed to take their BP daily and record BP readings in a paper log. During the pretelehealth period, adherence with at least one clinic visit was 48.5% for non-Hispanic Black patients (76.3% of sample) and 73.1% for non-Hispanic White patients (racial gap 24.6%). In the posttelehealth period, 76.4% of non-Hispanic Black patients and 76.7% of non-Hispanic White patients attended a telehealth visit (racial gap 0.4%). In the multivariable model, the overall improvement in adherence with BP checks was associated with increased participation by non-Hispanic Black patients in telehealth visits compared with standard clinic visits (P<.001). Postpartum readmission rates within 6 weeks of delivery were similar between the pretelehealth and posttelehealth periods overall (17.8% vs 17.4%, respectively) and when stratified by race.

Single-Arm Studies of Home Blood Pressure Monitoring

Eight studies offered home BP monitoring to all individuals who met specific inclusion criteria (Table 2).^{23,27–29,35,36,38,40} Four of the studies described themselves as feasibility studies that assessed the practicality of home BP monitoring in a small group of participants (See Appendix 3, available online athttps://links.lww.com/AOG/D229, for a detailed summary of these studies).^23,27,35,38 Among those who chose to participate in feasibility studies of home BP monitoring, 83.3–87.0% reported that they were satisfied with their postpartum care.

The remaining four studies included three implementation studies^29,36,40 and one quality-improvement study²⁸ of incorporation of home BP monitoring into routine clinical practice. Overall, these studies found that home BP monitoring improves reporting of BP measurements, allows ascertainment of elevated BPs, and possibly facilitates initiation of needed urgent or emergent care. In two studies, 95.5%⁴⁰ and 96.5%³⁶ of patients submitted at least one BP measurement (and ascertainment of BP was similar by race⁴⁰), and the American College of Obstetricians and Gynecologists’ recommendation for a BP measurement on postpartum day 3 or 4 and again between days 7 and 10 was achieved by 63.3% and 84.7% of participants, respectively.^36,40 Home BP monitoring allowed 43.2% of patients to replace an in-office BP check at 1 week.²⁹ In one study, 16.1% of patients had antihypertensive medication initiated based on home BP monitoring³⁶; in another study, home BP monitoring was used to initiate or titrate antihypertensive medications in 41.8% of patients.²⁹ Most patients continued with the home BP–monitoring program through the study: 94.9% at 2 weeks, 83.1% at 3 weeks, 73.8% at 4 weeks, and 67.5% for 5 weeks or more.²⁹ In one study,²⁹ 94.0% of survey respondents reported that they were satisfied with the program; in another study,³⁶ 91.9% stated that they would recommend the program to others, 94.9% said text messages were easy to receive, and 92.9% said the program helped them pay attention to their BP. In one single-arm study, all postpartum individuals were discharged with a BP cuff and contacted by telephone 1 week postpartum. Among those without a prior HDP diagnosis, 7.7% were referred to the emergency department for evaluation of a severe-range BP and 3.6% were diagnosed with de novo severe postpartum preeclampsia.²⁸

Limitations of the Evidence

The evidence base is limited, consisting of only five comparative studies, none of which was powered to detect differences in patient outcomes. Specifically, evidence for the effectiveness of home BP monitoring compared with usual clinic-based care was sparse (two RCTs and one nonrandomized comparative study). We found insufficient evidence to assess whether home BP monitoring reduces severe maternal morbidity or mortality or reduces existing racial disparities in morbidity and mortality or to assess for potential harms associated with home monitoring. A single, as yet unpublished, RCT with partial results reported onClinicalTrials.gov (NCT 03728790) addresses the comparative effectiveness of two home BP–monitoring strategies. The sparse data and variably reported outcomes preclude a pooled summary of overall effects.

DISCUSSION

Home BP monitoring probably improves overall BP ascertainment (determination of BP measures) in the early postpartum period, and patients appear to be generally satisfied with home BP–monitoring management. Home BP monitoring may reduce hypertension-related hospital admissions and probably reduces disparities in BP ascertainment between Black and non-Black patients by compensating for lower rates of postpartum office visit attendance among Black patients.⁴²

Improved ascertainment and monitoring of postpartum BP is necessary to identify hypertension but is by no means sufficient to reduce maternal morbidity and mortality, unless combined with ready access to a health care professional for medication initiation and titration and further urgent or emergent evaluation and treatment if needed.

Early recognition of hypertension, combined with effective communication and initiation of antihypertensive treatment, has the potential to decrease the need for rehospitalization. However, the effects on health care utilization may be complex because home BP monitoring will prompt referral of some individuals for urgent or emergent evaluation, which may result in an increase in emergency department visits and rehospitalization.

All effectiveness and comparative effectiveness studies included in this review enrolled individuals treated in urban, academic medical centers, which may not be generalizable to all communities and settings. In rural areas of the United States with limited internet access⁴³ and few specialist health care professionals,⁴⁴ implementation of home BP monitoring may be particularly helpful, albeit challenging.

Large pragmatic trials enrolling participants with antepartum HDP diagnoses, adequately powered to detect clinical outcomes, are needed. With one exception,²⁸ studies enrolled participants with an antepartum HDP diagnosis. We agree with the draft recommendation of the related AHRQ review, “Screening for Hypertensive Disorders in Pregnancy: An Evidence Update for the U.S. Preventive Services Task Force,”⁴⁵ that more research is needed regarding the comparative benefits and harms of universal home BP screening to detect new-onset (de novo) postpartum preeclampsia in individuals without an HDP diagnosis.

Analyses of postpartum BP trajectories have found that postpartum BP peaks on postpartum days 5–7, with a rapid decrease from postpartum day 7 until postpartum day 14. Home BP monitoring has been used to facilitate research on postpartum antihypertensive medications in two recent RCTs—a trial of furosemide compared with placebo⁴⁶ and a trial of combined hydrochlorothiazide and lisinopril compared with nifedipine.⁴⁷ A substantial percentage of patients with HDP continued to require antihypertensive medications to maintain BP lower than 140/90 mm Hg at 6 weeks, including 32.6% of patients with preeclampsia with severe features, 23.8% of patients with preeclampsia without severe features, and 16.1% of patients with gestational hypertension.⁴⁸ Thus, many individuals may benefit from longer-duration home BP monitoring with medication titration and lifestyle educational interventions.⁴⁹ Notably, one trial found that home BP monitoring resulted in small, sustained reductions of diastolic BP up to 3–4 years postpartum.²⁶ If confirmed,⁵⁰ these findings suggest that better control of postpartum hypertension may potentially reduce longer-term cardiovascular risk.

Home BP–monitoring interventions consist of multiple potential components, including education, availability of a BP cuff, communications technology (ie, text messaging, smartphone or tablet, web-based comprehensive telehealth), and initiation or titration of antihypertensive medications. Preliminary results from a single comparative effectiveness trial suggest that web-based communication of BPs or a scheduled audio-only telehealth visit is more effective than provision of discharge education, a prescription for a BP cuff, and instructions to keep a paper log.³⁹

The Heart Safe Motherhood program, which has been evaluated in both an RCT and single- and multi-center cohorts, relies on relatively “low-tech” text messaging. The COVID-19 pandemic has greatly increased the use of telemedicine. There is a need for further research to compare the advantages and disadvantages of available communication technologies. Clinicians and health systems providing care to postpartum individuals will need to create care pathways and quality-improvement initiatives that incorporate home BP monitoring, identification of postpartum hypertension, treatment, and coordination of care for these individuals at risk from worsening or newly elevated BPs.³⁷ Tailoring these interventions to individuals’ comfort with technology, access to digital tools, and preferences may be a helpful step in improving uptake and adherence.

Given the very large disparities in health care and health outcomes related to HDP that adversely affect Black individuals and individuals in other marginalized groups in the United States, including American Indian and Alaska Natives,⁸ people living in rural areas, and those living on low-incomes, future research should focus on reducing these disparities. Such research (incorporating both qualitative and quantitative methods) will require community partnerships and diverse multidisciplinary research teams outside of urban, academic medical centers.⁵¹

Home BP monitoring in postpartum individuals probably improves ascertainment of BP, allowing early recognition of hypertension, and probably compensates for racial disparities in office-based follow-up. There is insufficient evidence to conclude whether home BP monitoring reduces severe maternal morbidity or mortality or racial disparities in clinical outcomes.

REFERENCES

Supplemental Digital Content

• AOG_142_2_2023_05_27_STEELE_23-630_SDC1.pdf; [PDF] (270 KB)
• AOG_142_2_2023_05_27_TPRSTEELE_23-630_SDC2.pdf; [PDF] (526 KB)