Population-based data have demonstrated a doubling of the risk of stroke associated with hypertensive diseases of pregnancy between 1994 and 2011.1 Severe range blood pressure (BP), and in particular systolic blood pressure of 160 mm Hg or higher, may be associated with pregnancy-associated stroke.2,3 Data from the Hospital Corporation of America demonstrated decreased risk for fatal hypertension-related intracranial hemorrhage associated with protocolized, timely administration of antihypertensive medication.4 Goals of antihypertensive therapy also include prevention of renal failure and injury and cardiovascular morbidity (specifically congestive heart failure and myocardial ischemia).5 The American College of Obstetricians and Gynecologists' (ACOG) Task Force on Hypertension in Pregnancy supports administration of antihypertensive agents for preeclampsia with severe features with sustained BP of 160 mm Hg or 110 mm Hg systolic and diastolic, respectively, with additional recommendations for treatment of postpartum hypertension and women with chronic hypertension. First-line agents for treatment of acute hypertension include intravenous (IV) labetalol, IV hydralazine, and oral nifedipine.5 The National Partnership for Maternal Safety has disseminated a bundle to aid in timely treatment of severe hypertension.6
In the setting of these recommendations and evolving evidence demonstrating the benefit of treating pregnancy-associated hypertension, there is limited research on how practice patterns are changing and specifically to what degree antihypertensive agents are being used to treat hypertensive episodes. How temporal trends, patient factors, and hospital-level effects may affect use of hypertensive agents is not well characterized. Additionally, there are little data on the degree to which treatment with antihypertensive agents is associated with the improved outcomes.
Given that hypertension is a leading cause of maternal mortality and severe morbidity7 and that there are knowledge gaps in how management is evolving, the primary objective of this study was to characterize antihypertensive use during delivery hospitalizations complicated by preeclampsia. We hypothesized that use of antihypertensive agents has increased concurrent with expanding evidence and supportive clinical recommendations. A secondary objective of this analysis was to characterize risk for pregnancy-associated stroke for women with a preeclampsia diagnosis.
MATERIALS AND METHODS
The Perspective database was used for this retrospective cohort analysis. Perspective is maintained by Premier Incorporated (Charlotte, North Carolina) and includes patient demographics, hospital characteristics, and discharge diagnosis codes as well as medications and devices administered during acute care hospitalizations. Perspective reports 100% of hospitalizations for individual hospitals. Ninety-five quality assurance and validation checks are performed on data before being used for research.8 Perspective is routinely used for research on trends on medications and device use during delivery hospitalizations.9–15 The discharges included in the Perspective database account for approximately 15% of all inpatient hospital stays annually in the entire United States. The Columbia University institutional review board deemed the study exempt given that all data are deidentified.
Women included in this analysis were admitted for a delivery hospitalization with an associated preeclampsia diagnosis from January 2006 through March 2015. Patients with preeclampsia were identified based on International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes and subcategorized as mild, severe, and superimposed (ICD-9-CM codes 642.4x, 642.5x, and 642.7x, respectively). Reports of sensitivity for ICD-9-CM codes for preeclampsia range from 69% to 84%.16–18 Delivery hospitalizations were identified based on billing and procedure codes that ascertain more than 95% of deliveries.19 Among patients with preeclampsia, we analyzed whether patients received the following antihypertensive medications: oral labetalol, IV labetalol, nifedipine, and IV hydralazine. We evaluated temporal trends in rates of individual antihypertensive use within this cohort 1) for all patients with preeclampsia and 2) individually for mild, severe, or superimposed preeclampsia. Temporal trends are reported through 2014 given that only data for the first quarter of 2015 were available.
Univariate associations between clinical, hospital, and demographic variables and receipt of antihypertensive medications were evaluated using the χ2 test. Adjusted risk ratios (RRs) for receipt of any antihypertensive medication with 95% CIs as measures of effect accounting for demographic, hospital, and preeclampsia diagnosis were derived from fitting a log-linear generalized estimating equations regression model that accounts for the effect of patients clustered within hospitals. Patient demographic characteristics included year of discharge, maternal age, marital status, race (white, black, other), and payer status. Hospital characteristics included bed size, geographic region, teaching status, and urban or rural location. Specific preeclampsia diagnosis (severe, mild, or superimposed) was included in the model. We additionally performed a sensitivity analysis to address potential confounding given that the hospitals included in Perspective change from year to year and temporal trends could be secondary to inclusion of specific hospitals; to address this changing sampling frame, we repeated both the adjusted and unadjusted analysis restricted to hospitals that contributed data for the entire study period. Given maternal disparities and increased risk for maternal mortality among black women,7 we performed an additional analysis stratified for receipt of any antihypertensive by race (black, white, other) over the study period. Given that uptake of clinical recommendations may be different for teaching compared with nonteaching centers, we also report on temporal trends in receipt of antihypertensive agents for preeclamptics based on this factor.
To determine whether risk for pregnancy-associated stroke for patients with preeclampsia changed over the study period, we analyzed the number of strokes that occurred restricted to ICD-9-CM codes with high sensitivity ascertained in nonobstetric validation studies (ICD-9-CM 430, 431, 434.x1, and 436) as well as ICD-9-CM code 674.0, which we have found to be highly sensitive in pregnancy-related stroke in our institution.20–24 Because pregnancy-associated stroke is a relatively rare event occurring during approximately 34 per 100,000 pregnancies,25,26 we divided the study into three time periods: 2006–2008, 2009–2011, and 2012–2014 and compared risk using the χ2 test with the numerator as stroke cases and the denominator as all delivery hospitalizations. Risk for stroke in 2015 was not reported given that only the first quarter was available. We performed this analysis for all women with preeclampsia and women with a diagnosis of severe preeclampsia alone. Lastly, to evaluate variation in use of antihypertensive agents in the setting of severe preeclampsia across hospitals, we restricted the analysis to centers that treated 50 or more patients with severe preeclampsia over the study period and calculated hospital-level rates of antihypertensive use for this subset of patients. All analyses were performed with SAS 9.4.
Overall 239,454 patients with preeclampsia were included in the analysis. Of these, 126,595 patients had mild, 31,628 patients had superimposed, and 81,231 patients had severe preeclampsia (Table 1). The proportion of patients with severe and superimposed relative to mild preeclampsia increased over the study period with severe preeclampsia accounting for 36.9% of cases in 2015 compared with 30.9% of cases in 2006. In 2006, 10.3% and 30.9% of patients received diagnoses of superimposed and severe preeclampsia, respectively, compared with 15.7% and 36.9% in 2015. Superimposed preeclampsia was more common in the South, among black women, women 35 years of age or older, and in teaching compared with nonteaching hospitals. Severe preeclampsia was also more common in teaching compared with nonteaching hospitals as well as hospitals with larger bed sizes (more than 600 beds).
Over the course of the study, the proportion of patients with preeclampsia receiving any antihypertensive medication rose from 37.8% of patients in 2006 to 49.4% in 2015 (P<.01) (Table 1). Figure 1 shows receipt of individual hypertensive agents (IV labetalol, oral labetalol, nifedipine, hydralazine) by year for all patients with preeclampsia and by specific preeclampsia diagnosis: mild, superimposed, or severe. For all patients with preeclampsia, receipt of four individual medications increased over the study period from 2006 to 2014 (Fig. 1A): oral labetalol from 20.3% to 31.4%, IV labetalol from 13.3% to 21.4%, hydralazine from 12.8% to 16.9%, nifedipine from 15.0% to 18.2%, and more than one medication from 16.5% to 25.8%. Patients with severe preeclampsia (Fig. 1B) also received medications more often over the study period: between 2006 and 2014, use of IV labetalol increased 11.1%, oral labetalol 10.1%, hydralazine 4.0%, nifedipine 3.9%, and more than one agent 10.7% (absolute increases in administration). For superimposed preeclampsia (Fig. 1C), use of IV labetalol increased 6.8%, oral labetalol 11.1%, hydralazine 2.3%, nifedipine 3.5%, and more than one agent 7.8% (absolute increases in administration). Increases in antihypertensive use among mild preeclamptic women were smaller (Fig. 1D) with use of IV labetalol increasing 3.0%, oral labetalol 7.0%, hydralazine 1.0%, nifedipine 0.9%, and more than one agent 3.9% over the same period (absolute increases in administration). Table 2 demonstrates the number of patients each year with each preeclampsia diagnosis receiving each medication. Individual hospital rates of antihypertensive administration to patients with severe preeclampsia differed significantly (Fig. 2). The median hospital-level rate of antihypertensive administration to patients with severe preeclampsia was 60.0% with the 25th and 75th percentiles 52.7% and 67.9%, respectively.
In the adjusted analysis, the likelihood of antihypertensive use increased over the study period with an adjusted RR of 1.19 and 95% CI in 2015 with 2006 as a reference (Table 3). Severe preeclampsia and superimposed preeclampsia were associated with the largest increase in likelihood of antihypertensive receipt (adjusted RR 2.19, 95% CI 2.12–2.26; adjusted RR 2.32, 95% CI 2.23–2.41) with mild preeclampsia as a reference, respectively. Other significant factors included black compared with white race (adjusted RR 1.30, 95% CI 1.27–1.33) and maternal age 35 years or older compared with age 18–24 years (adjusted RR 1.17, 95% CI 1.15–1.19). For the sensitivity analysis, when the cohort was restricted only to centers contributing data for the entire study period, temporal trends and adjusted analyses demonstrated similar results (not shown). In evaluating antihypertensive agents by race, black women were most likely to receive medication with receipt of any hypertensive increasing from 53.1% to 60.9% over the study period compared with 33.6%–45.6% for white women (Fig. 3). Rates of antihypertensive receipt were similar at teaching and nonteaching hospitals (Fig. 4).
Risk for pregnancy-associated stroke over the study period for women with severe preeclampsia and any preeclampsia decreased (Fig. 5). For severe preeclampsia, the risk for stroke was highest from 2006 to 2008 (13.5 cases/10,000 deliveries, n=27) and decreased to 9.7 per 10,000 deliveries in 2009–2011 (n=25) and 6.0 per 10,000 deliveries in 2012–2014 (n=20) (P=.02) (Table 4). For all patients with preeclampsia, the risk for stroke decreased from 6.6 to 5.6 to 3.5 per 10,000 deliveries, respectively, over the same study period (P=.02).
This analysis demonstrated increasing use of antihypertensive agents during delivery hospitalizations complicated by preeclampsia. The rates of use of IV labetalol, oral labetalol, nifedipine, and hydralazine as well as use of multiple agents all increased for women with mild, superimposed, and severe preeclampsia over the study period. These data support that research evidence supporting the benefit of antihypertensive use4 may be becoming integrated into practice as patients benefitting from antihypertensive management are increasingly identified by health care providers and hospital protocols. Given that ACOG’s Task Force on Hypertension report was released in November 2013 and recommendations from the National Partnership were released this year, further changes in clinical practice may be occurring.5,6 Although ACOG’s Task Force report likely affected trends late in the study period, these data support that antihypertensive use was already increasing before release of ACOG’s recommendations. That use of both IV and oral medications increased suggests that medical management of hypertension in patients with preeclampsia is changing comprehensively. Although misclassification between severe and mild preeclampsia is a potential concern, increased use of antihypertensive agents for women with mild preeclampsia may represent health care providers treating more lower range BPs more frequently, particularly with oral agents before discharge.
Important findings from our analysis include that, particularly for women with severe preeclampsia, health care providers use IV and oral labetalol preferentially. With expanded use of these medications, appropriate patient selection and use of other agents for women with conditions such as asthma, ischemic heart disease, and heart failure5 may maximize benefit and minimize risk. In this analysis, the effect of individual hospitals was an important determinant in medication receipt for severe preeclampsia. Rates of medication receipt varied significantly and further investigation is indicated to what degree protocol, guideline, and bundle adoption can facilitate uniformly high-quality care that minimizes variation.
In interpreting this analysis, there are several important limitations to consider. First, because this study was limited to administrative data, we were not able to obtain detailed information on how hospital factors such as protocols, guidelines, and bundle implementation may have affected management of preeclampsia and outcomes. Second, our ability to characterize patient care is limited given that we do not have patient data on vital signs and time to hypertensive administration. Although a drug may have been ordered and administered, we cannot determine whether care occurred in a timely, optimal fashion and whether treatment in a specific time interval improved outcomes. For nifedipine, we are not able to determine whether medication is being administered acutely for severe range hypertension or preemptively for longer term BP control. Furthermore, because we are not able to review BP, we are not able to evaluate risk for hypotension with antihypertensive administration and potential effects on the fetus nor are we able to determine specific BP readings before stroke. Third, we cannot establish a causal relationship between antihypertensive administration (or lack thereof) and risk for stroke. Although data from the Hospital Corporation of America support a relationship between timely administration of antihypertensive agents and decreased risk, we are not able to determine whether patients with stroke received hypertensive management before their event per the recommendations of ACOG and others and we are not able to evaluate hospital-level implementation of protocols.5,6,27 Fourth, because we do not have outpatient data, we are unable to evaluate the degree to which prenatal outpatient management and use of antihypertensive agents may have affected use of these medications in the hospital. Fifth, because we do not have BP information, we do not have a true denominator for how many patients should have been treated. Sixth, because many patients were coded with nonspecific pregnancy-related stroke codes, for a significant proportion of the study, it is unclear whether women were diagnosed with a hemorrhagic or thrombotic stroke. Seventh, because discharge diagnoses are used, this study does not have information on whether stroke occurred antepartum, intrapartum, or postpartum during a delivery hospitalization. Eight, because ascertainment of obesity is extremely poor in administrative data, this factor was not included in our analysis. Ninth, use of multiple agents may represent sequential medication administration (for example, a new agent being utilized after a suboptimal response from a prior agent) or use of multiple concurrent agents (such as a longer acting medication for maintenance along with an acute agent for severe hypertension); this analysis cannot differentiate these clinical scenarios. Tenth, in some cases, hydralazine may be administered orally. In our query, in the vast majority of instances, hydralazine was administered intravenously; however, there were a few instances in which route of administration was not documented specifically. For this reason, our reported rates of IV hydralazine administration could be slightly overestimated.
Strengths of the study include a large patient population powered to assess rare outcomes such as stroke, a long study period, geographically and clinically diverse hospital settings, and the ability to query use of specific antihypertensive medications.
In summary, this analysis found use of antihypertensive agents increased among a population of women with increasing likelihood of severe and superimposed preeclampsia diagnoses. Use of antihypertensive agents increased for diagnoses of mild, moderate, and severe hypertension. This trend of increased antihypertensive use was associated with a decrease in risk for pregnancy-associated stroke over the study period.
1. Leffert LR, Clancy CR, Bateman BT, Bryant AS, Kuklina EV. Hypertensive disorders and pregnancy-related stroke: frequency, trends, risk factors, and outcomes. Obstet Gynecol 2015;125:124–31.
2. Martin JN Jr, Thigpen BD, Moore RC, Rose CH, Cushman J, May W. Stroke and severe preeclampsia and eclampsia: a paradigm shift focusing on systolic blood pressure. Obstet Gynecol 2005;105:246–54.
3. Martin JN Jr. Severe systolic hypertension and the search for safer motherhood. Semin Perinatol 2016;40:119–23.
4. Clark SL, Christmas JT, Frye DR, Meyers JA, Perlin JB. Maternal mortality in the United States: predictability and the impact of protocols on fatal postcesarean pulmonary embolism and hypertension-related intracranial hemorrhage. Am J Obstet Gynecol 2014;211:32.e1–9.
5. American College of Obstetricians and Gynecologists; Task Force on Hypertension in Pregnancy. Hypertension in pregnancy. Report of the American College of Obstetricians and Gynecologists' Task Force on Hypertension in Pregnancy. Obstet Gynecol 2013;122:1122–31.
6. Bernstein PS, Martin JN Jr, Barton JR, Shields LE, Druzin ML, Scavone BM, et al. National Partnership for Maternal Safety: consensus bundle on severe hypertension during pregnancy and the postpartum period. Obstet Gynecol 2017;130:347–57.
7. Creanga AA, Berg CJ, Syverson C, Seed K, Bruce FC, Callaghan WM. Pregnancy-related mortality in the United States, 2006–2010. Obstet Gynecol 2015;125:5–12.
8. Stulberg JJ, Delaney CP, Neuhauser DV, Aron DC, Fu P, Koroukian SM. Adherence to surgical care improvement project measures and the association with postoperative infections. JAMA 2010;303:2479–85.
9. Fang MC, Maselli J, Lurie JD, Lindenauer PK, Pekow PS, Auerbach AD. Use and outcomes of venous thromboembolism prophylaxis after spinal fusion surgery. J Thromb Haemost 2011;9:1318–25.
10. Ritch JM, Kim JH, Lewin SN, Burke WM, Sun X, Herzog TJ, et al. Venous thromboembolism and use of prophylaxis among women undergoing laparoscopic hysterectomy. Obstet Gynecol 2011;117:1367–74.
11. Wright JD, Lewin SN, Shah M, Burke WM, Lee SM, Sun X, et al. Quality of venous thromboembolism prophylaxis in patients undergoing oncologic surgery. Ann Surg 2011;253:1140–6.
12. Zacharia BE, Youngerman BE, Bruce SS, Hershman DL, Neugut AI, Bruce JN, et al. Quality of postoperative venous thromboembolism prophylaxis in neuro-oncologic surgery. Neurosurgery 2017;80:73–81.
13. Prabhakaran S, Herbers P, Khoury J, Adeoye O, Khatri P, Ferioli S, et al. Is prophylactic anticoagulation for deep venous thrombosis common practice after intracerebral hemorrhage? Stroke 2015;46:369–75.
14. Kulik A, Rassen JA, Myers J, Schneeweiss S, Gagne J, Polinski JM, et al. Comparative effectiveness of preventative therapy for venous thromboembolism after coronary artery bypass graft surgery. Circ Cardiovasc Interv 2012;5:590–6.
15. Shaw AD, Bagshaw SM, Goldstein SL, Scherer LA, Duan M, Schermer CR, et al. Major complications, mortality, and resource utilization after open abdominal surgery: 0.9% saline compared to Plasma-Lyte. Ann Surg 2012;255:821–9.
16. Klemmensen AK, Olsen SF, Osterdal ML, Tabor A. Validity of preeclampsia-related diagnoses recorded in a national hospital registry and in a postpartum interview of the women. Am J Epidemiol 2007;166:117–24.
17. Yasmeen S, Romano PS, Schembri ME, Keyzer JM, Gilbert WM. Accuracy of obstetric diagnoses and procedures in hospital discharge data. Am J Obstet Gynecol 2006;194:992–1001.
18. Lydon-Rochelle MT, Holt VL, Nelson JC, Cárdenas V, Gardella C, Easterling TR, et al. Accuracy of reporting maternal in-hospital diagnoses and intrapartum procedures in Washington State linked birth records. Paediatr Perinatal Epidemiol 2005;19:460–71.
19. Kuklina EV, Whiteman MK, Hillis SD, Jameieson DJ, Meikle SF, Posner SF. An enhanced method for identifying obstetric deliveries: implications for estimating maternal morbidity. Matern Child Health J 2008;12:469–77.
20. Jones SA, Gottesman RF, Shahar E, Wruck L, Rosamond WD. Validity of hospital discharge diagnosis codes for stroke: the Atherosclerosis Risk in Communities Study. Stroke 2014;45:3219–25.
21. Roumie CL, Mitchel E, Gideon PS, Varas-Lorenzo C, Castellsague J, Griffin MR. Validation of ICD-9 codes with a high positive predictive value for incident strokes resulting in hospitalization using Medicaid health data. Pharmacoepidemiol Drug Saf 2008;17:20–6.
22. Thigpen JL, Dillon C, Forster KB, Henault L, Quinn EK, Tripodis Y, et al. Validity of International Classification of Disease codes to identify ischemic stroke and intracranial hemorrhage among individuals with associated diagnosis of atrial fibrillation. Circ Cardiovasc Qual Outcomes 2015;8:8–14.
23. Miller EC, Gatollari HJ, Too G, Boehme AK, Leffert L, Elkind MS, et al. Risk of pregnancy-associated stroke across age groups in New York State. JAMA Neurol 2016;73:1461–7.
24. McCormick N, Bhole V, Lacaille D, Avina-Zubieta JA. Validity of diagnostic codes for acute stroke in administrative databases: a systematic review. PLoS One 2015;10:e0135834.
25. Moatti Z, Gupta M, Yadava R, Thamban S. A review of stroke and pregnancy: incidence, management and prevention. Eur J Obstet Gynecol Reprod Biol 2014;181:20–7.
26. James AH, Bushnell CD, Jamison MG, Myers ER. Incidence and risk factors for stroke in pregnancy and the puerperium. Obstet Gynecol 2005;106:509–16.
© 2018 by American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
27. Moroz LA, Simpson LL, Rochelson B. Management of severe hypertension in pregnancy. Semin Perinatol 2016;40:112–8.