Hypertens Pregnancy, 2014; 33(4): 476–487 ! Informa Healthcare USA, Inc. ISSN: 1064-1955 print / 1525-6065 online DOI: 10.3109/10641955.2014.946612
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
ORIGINAL ARTICLE
Measurement of aortic augmentation index in pregnant women with raised blood pressure and subsequent outcomes: a preliminary prospective cohort study Gail Fullerton,1 Michael A. Crilly,2 Sohinee Bhattacharya,2 and Peter J. Danielian1 1
Department of Obstetrics, Aberdeen Maternity Hospital, Foresterhill, Aberdeen, UK and 2 Division of Applied Health Sciences, University of Aberdeen, Foresterhill, Aberdeen, UK Objective: Preeclampsia is associated with arterial dysfunction and augmentation index (AIX%) is an established indicator of arterial dysfunction. Our aim was to investigate the relationship of AIX% with time-to-delivery and other outcomes in women admitted to an antenatal triage unit. Methods: We recruited 28 women with singleton pregnancies attending antenatal triage ward for assessment of hypertension. After 10 min rest, seated brachial blood pressure (Omron HEM-757) and AIX% (SphygmoCor applanation tonometry pulse wave analysis, PWA) were measured by a single investigator; other clinicians remained blinded to PWA results. Routine assessment included cardiotocography, urine analysis and blood tests. Subsequent outcomes were extracted from the obstetric records. Results: Mean AIX% was 19.7% (SD 11.5; range 4% to +36%), maternal age 31 years, gestation 37 weeks, brachial BP 145/95, proteinuria 39%. Nine women had preeclampsia at assessment and six subsequently developed preeclampsia. Median time-to-delivery was 10 d (IQR 1.6–25 d) and was shorter for AIX% 20% (median 8.9 versus 19.8 d). AIX% was higher with preeclampsia (24.0%; SD 9.5) versus gestational hypertension (15.2%; SD 12.4); absolute difference 8.8% (95%CI 0.1–17.5; p = 0.05). A one-point higher AIX% (adjusted for age, urate and gestation) was associated with 0.3 d (95%CI 0.5 to 0.0; p = 0.06) reduced time-to-delivery. A higher AIX% was also associated with induction for preeclampsia, severe preeclampsia, peripartum–antihypertensives and discharge-on-anti-hypertensives. Area under the curve (AUC) for AIX% predicting preeclampsia was 0.80 (95%CI 0.59–1.00; p = 0.04). Conclusion: AIX% is associated with time-to-delivery and other outcomes in pregnancy. Keywords Augmentation index, Applanation tonometry, Hypertension, Pregnancy, Preeclampsia, Pulse wave analysis.
Correspondence: Dr Michael A. Crilly, MD MPH MRCGP MFPHM, Senior Lecturer in Clinical Epidemiology, Division of Applied Health Sciences, University of Aberdeen, Polwarth Building, Foresterhill, Aberdeen AB25 2ZD, UK. Tel: +44(0)1224 437855. Fax: +44(0)1224 550925. E-mail: [email protected]
Augmentation index and preeclampsia
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
INTRODUCTION The major temporal haemodynamic changes that occur in normal pregnancy include an increase in arterial wall compliance, reduced arterial resistance and a reduction in the magnitude of arterial wave reflection within the arterial tree (1). This is in contrast with the widespread endothelial dysfunction and increased peripheral arterial resistance that is observed in pregnancies complicated by preeclampsia (2–4). Preeclampsia is an important indication for the induction of labour and a cause of premature delivery (2). Arterial wave reflection can be assessed non-invasively using pulse waveform analysis (PWA) combined with applanation tonometry at the radial artery. Using commercially available PWA devices, such as the SphygmoCor system, it is possible to rapidly assess central aortic pulse waves and pressures in routine clinical settings without the need for cardiac catheterisation (5–7). SphygmoCor PWA can be used to measure the augmentation index (AIX%), which depends upon the extent of arterial wave reflection (5). PWA is becoming an increasingly popular technique for assessing haemodynamics in pregnancy (8–16). The out-going systolic pulse wave generated by the heart in the conduit and peripheral arteries is also reflected back towards the heart where it can add to (‘‘augment’’) the central aortic pressure (5). The amount by which the central aortic pressure is increased by this phenomenon of wave reflection is the ‘‘augmentation pressure’’; the AIX% is the ‘‘augmentation pressure’’ (AP) expressed as a proportion of the central aortic ‘‘pulse pressure’’ (PP). AIX% represents a composite index of arterial function that is influenced by endothelial dysfunction, arterial compliance, peripheral arterial resistance and left ventricular ejection (5). Previous research in non-pregnant individuals has demonstrated that AIX% is closely associated with endothelial dysfunction (17,18). In non-pregnant populations, a higher AIX% has been shown to be associated with a higher rate of major adverse cardiovascular events such as myocardial infarction and stroke (19). Recent PWA studies in uncomplicated pregnancies have demonstrated that AIX% declines in the first half of pregnancy, reaches its nadir around midpregnancy and then rises again to early pregnancy levels at term (12,13). Compared to normotensive pregnancies AIX% is substantially higher in pregnancies complicated by hypertension (8–10,14,16). AIX% is also higher in preeclamptic pregnancies compared to those complicated by gestational hypertension (8,9,16). The majority of previous studies have been crosssectional, but a single prospective cohort study has demonstrated that a higher AIX%, measured at around 13 weeks gestation, is associated with an increased risk of subsequently developing preeclampsia (12). However, no previous studies have assessed the relationship of AIX% with subsequent outcomes in women with pregnancies complicated by hypertension. We wondered if the additional measurement of AIX% in women being assessed for hypertension in the second half of pregnancy might be helpful in identifying those women destined for a less benign clinical course. If this were the case then measuring AIX% might be useful in guiding the subsequent obstetric monitoring and intervention in pregnant women with hypertension. The aim of this preliminary prospective cohort study is to assess the relationship between obstetric outcomes and AIX% measured non-invasively
477
478
G. Fullerton et al.
in pregnant women admitted to a triage unit for hypertension in a routine healthcare setting.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
MATERIALS AND METHODS Women with a singleton pregnancy who attended the antenatal triage ward at Aberdeen Maternity Hospital (January to August 2012) with suspected hypertensive disorder were enrolled in this study. Women attended the antenatal triage ward for the assessment of high blood pressure and/or proteinuria following either: hospital antenatal clinic assessment, community midwife assessment, or after contacting the unit complaining of symptoms associated with preeclampsia (such as frontal headache, visual disturbance or epigastric pain). Routine clinical assessment of women admitted to the antenatal triage ward included aural temperature, brachial blood pressure, dipstick urine analysis (Siemens Multistix-8SG), venous blood sample (full blood count, urea, creatinine, liver function tests, uric acid) and cardiotocography. When proteinuria was present on dipstick testing, a midstream specimen of urine was sent for microscopy and culture. After resting seated for 10 min, all participants were assessed in the seated position by a single investigator (G. F.) on a single occasion. Right brachial BP was measured in triplicate using the Omron HEM-757 IntelliSense monitor with a standard cuff, or large sized cuff if required. The Omron M5-I /HEM-757 (Omron Healthcare Inc., Lake Forest, IL) is an automated oscillometric BP monitor, which has previously been independently validated against international criteria (20). The mean systolic and diastolic BP based on three readings was entered into the SphygmoCor device. Non-invasive PWA was also undertaken with the participant seated using the SphygmoCor device (SCOR v 8.0, AtCor Medical Pty Ltd., Sydney, Australia) with a hand-held high-fidelity tonometer (Millar Tonometer, Houston, TX) applanated (applied to flatten but not occlude) at the right radial artery. The SphygmoCor device captures waves over a 10-s period to produce an average pulse wave contour. At least two high quality PWA measurements were obtained from each woman. The quality of PWA was assessed visually on the laptop computer screen and numerically using the devices in-built percentage quality index score (QI%; device default setting based on: average pulse height, pulse height variation, diastolic variation, waveform shape deviation and maximum rate of pressure rise in the peripheral waveform). The SphygmoCor device calibrates the radial arterial waveform against the brachial blood pressure. The central aortic pulse wave is derived using a validated ‘‘generalised transfer function’’ which has previously been validated against simultaneous central aortic pressures obtained invasively by direct catheterisation (6,7). We have previously demonstrated the high level of repeatability of applanation tonometry PWA (21). All women underwent study assessment on the antenatal triage ward at Aberdeen Maternity Hospital between 10 a.m. and 10 p.m. Pulse wave velocity (PWV) was not measured. At initial study assessment information was also obtained concerning age, ethnicity, gestation, previous medical history, family history, current medication, current smoking status, number of previous deliveries 24 weeks and body mass index (BMI) recorded at the first antenatal booking visit. Laboratory results for full blood count, urea, creatinine, liver function and uric acid were
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
measured on admission to the antenatal triage ward. The subsequent treatment plan at ‘‘antenatal triage ward admission’’ was determined by medical staff not affiliated with the study who remained blinded to all of the PWA results throughout the study. The investigator (G. F.) was not involved in the clinical management of the women during the ‘‘triage ward admission’’, but she was involved in the clinical management of a small number of the women during subsequent admissions. Post-partum the obstetric case notes were reviewed by an obstetrician (G. F.) to determine subsequent use of peripartum (delivery ± 48 h) anti-hypertensive therapy, discharge home on anti-hypertensives, indications for induction of labour, mode of delivery, birth weight, major complications of pregnancy and the use of intravenous anti-hypertensives/ magnesium sulphate during labour. The highest brachial blood pressure and level of proteinuria during the peripartum period was recorded. Statistical Analysis The hypertensive status of women in the study was categorised according to International Society for the Study of Hypertension in Pregnancy (ISSHP) criteria as: normotensive if brachial BP5140/90 mmHg; gestational hypertension if BP 140/90 mmHg without proteinuria; and preeclamptic if BP 140/ 90 mmHg with proteinuria 1 + in the absence of a urinary tract infection (22). ‘‘Induction for preeclampsia’’ was defined as the use of vaginal prostaglandins, artificial rupture of membranes (ARM) and intravenous oxytocin in women who were not in labour with a brachial BP 140/90 and 1 + proteinuria on dipstick analysis. AIX% has a positive value if augmentation by the reflected wave occurs before the attainment of peak systolic pressure and a negative value if augmentation occurs after the attainment of peak systolic pressure (5). Since AIX% in an individual varies with heart rate it is commonly standardised to a heart rate of 75 beats-per-minute and this standardised AIX% (produced by the SphygmoCor device) was used in all analyses (23). Our pre-specified primary outcome was time-to-delivery (the SphygmoCor device automatically records that date and time of assessment and the maternity records the date and time of delivery). Secondary outcomes included peripartum (delivery ± 48 h) anti-hypertensive therapy, discharge home on anti-hypertensive therapy, induction for preeclampsia and the development of ‘‘severe’’ preeclampsia (defined as a brachial BP 160/ 110 mmHg, or BP 140/90 plus proteinuria 3+) (3). Continuous data approximating a normal distribution are presented as means (standard deviation) and skewed data are presented as medians (interquartile range, IQR). Mean arterial pressure (MAP, mmHg) was calculated as 1/3 systolic + 2/ 3 diastolic brachial blood pressure. Adjusted and unadjusted differences in mean AIX% were calculated and statistical precision was indicated using 95% confidence intervals (95%CI). Adjusted differences in mean AIX% were obtained using multiple linear regression (MLR) with forward selection of variables based on a probability for F of p = 0.08 for entry and p = 0.10 for removal. Statistical analysis was undertaken using IBM SPSS Statistics (version 20, IBM Corp., Armonk, NY). All p values are two-sided. All women provided written informed consent. The study adhered to the Helsinki Declaration and was approved by the North of Scotland Research Ethics Committee (14 November 2011; reference 11/NS/0039).
479
480
G. Fullerton et al.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
RESULTS A total of 31 women admitted to the antenatal triage ward for the assessment of hypertensive disorder in pregnancy were invited to participate in this study. A satisfactory PWA trace could not be obtained for one woman, one woman was excluded because of vasculitis and one woman declined to participate. The following analysis relates to the remaining 28 women. All were singleton pregnancies. No women were in a labour at the time of study assessment and all women with proteinuria ( 1+) had negative urinary microscopy and culture. AIX% ranged from 4% to +36% with a mean of 19.7 (SD 11.5) and a mean quality index (QI%) for PWA of 70% (SD 14). Women with an AIX% 20% had a higher brachial blood pressure, higher BMI at booking, higher urate level and were more likely to be nulliparous (Table 1). At initial study assessment, mean maternal age was 31 years, gestation 37 weeks (57% less than 38 weeks gestation), brachial BP 145/95 mmHg and 39% had proteinuria ( 1+) on urinary analysis (Table 1). The majority of women were nulliparous (75%) and Caucasian (93%). Only two women (7%) admitted to being current smokers and 24 (86%) reported never having smoked; two women were ex-smokers. Four women (14%) had a previous history of gestational hypertension and one woman had a prior history of essential hypertension. No women had a previous history of preeclampsia. Nine (32%) women reported a frontal headache at the time of assessment. One woman had multi-morbidity with rheumatoid arthritis, pernicious anaemia and hyperthyroidism (treated with sulphasalazine, prednisolone and vitamin B12). Table 1. Characteristics of women at initial study assessment. Augmentation index (AIX)
Mean maternal age (SD), years Mean gestation (SD), weeks Mean BMI at booking (SD), kg/m2 Mean brachial systolic BP, mmHg Mean brachial diastolic BP, mmHg Mean arterial pressure (SD), mmHg Nulliparous Brachial BP 140/90 mmHg Proteinuria ISSHP status at study assessment: Normotensive Gestational hypertension Preeclampsia Mean uric acid (SD), mmol/L Mean total bilirubin (SD), mmol/L Mean alanine-aminotransferase, IU/L Mean c-glutamyltransferase (SD), IU/L Mean creatinine (SD), mmol/L Mean platelets (SD), 109/L
AIX 20%
AIX520%
All women
(n = 17)
(n = 11)
(N = 28)
30 37.0 29 146 96 115 15 16 7
(5) (3.3) (5) (16) (8) (10) 88% 94% 41%
32 36.7 27 142 90 106 6 7 4
(4) (3.4) (5) (16) (11) (12) 55% 64% 36%
31 36.9 28 145 94 111 21 23 11
(4) (3.3) (5) (16) (10) (11) 75% 82% 39%
1 9 7 345 6 11 14 53 262
6% 53% 41% (44) (2) (4) (12) (5) (47)
4 5 2 285 7 14 10 50 237
36% 46% 18% (48) (2) (4) (3) (11) (53)
5 14 9 321 7 12 13 52 252
18% 50% 32% (53) (2) (4) (10) (8) (50)
All values are numbers (percentages) unless otherwise indicated. SD, standard deviation; BMI, body mass index; BP, blood pressure; ISSHP, International Society for the Study of Hypertension in Pregnancy.
Augmentation index and preeclampsia
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Obstetric Outcomes Labour was induced for 21 (75%) women among whom 10 were induced for preeclampsia. Twelve women (46%) had a spontaneous vaginal delivery, 10 (38%) caesarean section and 6 (23%) instrumental delivery. One woman had premature rupture of membranes at 28 weeks and was delivered by caesarean section; one woman had a major post-partum haemorrhage; and one woman had a caesarean section for foetal bradycardia during labour. Two women delivered babies with a birth weight below the 3rd Scottish centile (standardised for gestation and sex) (24). No other major complication of pregnancy occurred and no women had seizures. Time-to-Delivery The median time-to-delivery after study assessment was 10 d (IQR 1.6–25 d) and ranged from 0.5 to 61 d. Women with an AIX% 20% had a shorter timeto-delivery (median 8.9 d) compared to women with an AIX%520% (median 19.8 d). Unadjusted and adjusted regression coefficients in relation to time-todelivery are shown in Table 2. On unadjusted analysis a one-point higher AIX% at assessment was associated with 0.5 d (95%CI 1.1 to 0.1; p = 0.08) reduction in the time-to-delivery. Forward variable selection identified AIX%, gestation, urate and maternal age at initial assessment for inclusion in the regression model, which explained 82% of the variability in the time-todelivery (Table 2). On adjusted analysis a one-point higher AIX% at assessment was associated with 0.3 d (95%CI 0.5 to 0.0; p = 0.06) reduction in the time-to-delivery. Gestation at study assessment was the most important determinant of time-to-delivery with each additional week of gestation being associated with an adjusted 4.4 d (95%CI 3.5–5.3; p50.001) reduction in the time-to-delivery. Hypertensive Status At initial assessment, based on ‘‘study’’ brachial BP measurement and ISSHP criteria, nine (32%) women had preeclampsia, 14 (50%) women gestational hypertension and 5 (18%) were normotensive (Table 3). Four (14%) women were taking anti-hypertensive therapy at the time of study assessment and a further 8 (29%) women were commenced on anti-hypertensive therapy shortly after PWA-assessment. Sixteen women (57%) received peripartum antihypertensive therapy (with either labetalol or methyldopa); two of these Table 2. Factors associated with the primary outcome of time-to-delivery in days. Unadjusted coefficient Augmentation index, AIX% Gestation, weeks Uric acid, mmol/L Maternal age, years
0.50 4.50 0.13 0.64
Adjusted coefficienta 0.25 4.37 0.07 0.59
95%CI
95%CI
0.52 5.29 0.13 1.24
0.01 3.45 0.01 0.06
p 0.06 550.001 0.03 0.07
Summary of regression model: R 0.92, adjusted R2 0.82, p (ANOVA)550.001. Variables excluded by SPSS ‘‘forward selection’’: mean arterial pressure, proteinuria, nulliparous, BMI at booking, total bilirubin, alanine-aminotransferase, gamma-glutamyltransferase, creatinine, platelets.
a
481
482
G. Fullerton et al. Table 3. Hypertensive status at initial assessment and peripartum. Peripartum ISSHP status:
ISSHP status at assessment:
Normotensive
Gestational
Preeclampsia
Total nos (%)
Normotensive Gestational Preeclampsia Total nos (%)
1 1 0 2 (7%)
2 9 3 14 (50%)
2 4 6 12 (43%)
5 (18) 14 (50) 9 (32) 28 (100%)
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
ISSHP, International Society for the Study of Hypertension in Pregnancy.
women required intra-venous anti-hypertensive therapy during labour and one of these two women also received magnesium sulphate. Eleven (39%) women were discharged home on oral anti-hypertensive therapy. Change in hypertensive status between study assessment and the peripartum period (delivery ± 48 h) is shown in Table 3; the shaded boxes indicate the number of women whose hypertensive status remained unchanged. Hypertensive status remained unchanged for 16 (57%) women, worsened for 8 (29%) and improved for 4 (14%) women. Only one woman (study BP 135/89 with proteinuria and an elevated uric acid of 360 mmol/L) remained normotensive throughout pregnancy. Overall, 15 women met ISSHP criteria for preeclampsia during pregnancy, including 10 with ‘‘severe’’ preeclampsia (five at initial study assessment and five at peripartum); the remaining 12 women had gestational hypertension. AIX% was 24.0% (SD 9.5) in women with preeclampsia and 15.2% (SD 12.4) in women with gestational hypertension; an absolute difference of 8.8% (95%CI 0.1–17.5; p = 0.05). AIX% was highest among the five women with ‘‘severe’’ preeclampsia at initial study assessment (26.4%, SD 5.4). Secondary Outcomes The relationship of AIX% at initial study assessment with secondary outcomes is shown in Table 4. On adjusted analysis, AIX% was an absolute 10% higher in women with preeclampsia, those who require peripartum anti-hypertensive therapy and discharged home on anti-hypertensive therapy. Although not statistically significant at the 5% level, AIX% was also 8% higher among women induced for preeclampsia and among those with ‘‘severe’’ preeclampsia. Nine women met ISSHP criteria for preeclampsia at initial study assessment. Among the remaining 19 women free of preeclampsia at study assessment, 6 (32%) subsequently developed preeclampsia. Figure 1 shows the receiver operating characteristic (ROC) curve for these 19 women. The area under the curve (AUC) for the ability of AIX% 20% (i.e. above the mean AIX% value for the study) to predict the subsequent occurrence of preeclampsia was 0.80 (95%CI 0.59–1.00; p = 0.04).
DISCUSSION Among the women in this study being assessed for hypertension in pregnancy, a higher AIX% was independently associated with a shorter time-to-delivery after adjusting for maternal age, urate and gestation at baseline assessment
24.1 24.3 24.0 23.5 25.3
(10.9) (10.5) (9.5) (9.0) (9.2)
13.8 16.2 14.7 17.6 16.6
(10.0) (11.4) (12.0) (12.5) (11.7)
Mean AIX% (SD) Mean AIX% (SD)
Absent
10.3 8.1 9.3 5.9 8.8
Crude 10.4 9.6 9.9 8.0 7.5 b
b
b
a
a
Adjusted
Mean difference AIX
(2.4–18.2) (1.4–17.8) (2.1–17.6) ( 0.7 to 16.7) ( 1.3 to 16.2)
(95%CI)
0.01 0.02 0.02 0.07 0.09
p
0.55 0.52 0.56 0.49 0.47
R
0.24 0.21 0.31 0.18 0.16
Adj R2
0.01 0.02 0.01 0.03 0.04
p (ANOVA)
Summary regression model
Variables included by SPSS ‘‘forward selection’’: nulliparitya and mean arterial blood pressure b at initial study assessment. [Excluded variables (measured at initial study assessment): maternal age, gestation, proteinuria, uric acid, total bilirubin, alanine-aminotransferase, gammaglutamyltransferase, creatinine, platelets, body mass index at booking].
Peripartum anti-hypertensive therapy (16/12) Home on anti-hypertensive therapy (12/16) Preeclampsia (15/13) Severe preeclampsia (10/18) Induced for preeclampsia (10/18)
Secondary outcomes (present/absent)
Present
Table 4. Augmentation index (AIX%) at initial assessment and secondary outcomes.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
483
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
484
G. Fullerton et al.
Figure 1. Receiver operating characteristic (ROC) curve for augmentation index (AIX% 20%) and the occurrence of preeclampsia among women initially free of preeclampsia (n = 19).
(one-point increase in AIX% was associated with a 0.3 d [95%CI 0.5 to 0.0; p = 0.06] reduction in the time-to-delivery). AIX% was higher in women with preeclampsia compared to gestational hypertension. AIX% was also predictive of the subsequent development of preeclampsia after assessment. A higher AIX% was associated with induction of labour for preeclampsia, severe preeclampsia, peripartum-anti-hypertensive therapy and being discharged home on anti-hypertensive therapy. Study Strengths and Limitations Our study has some important strengths as it was undertaken by a single investigator in a routine healthcare setting using equipment that has previously been validated for the measurement of brachial BP (Omron M5-I/ HEM-757) (20) and PWA (SphygmoCor) (6,7). In order to ensure that women were haemodynamically stable at the time of assessment, PWA was undertaken after women had rested for at least 10 min. So as to avoid the knowledge of AIX% levels influencing clinical management, the medical staff providing obstetric care remained unaware of the PWA measurements, although the obstetrician (G. F.) undertaking PWA was involved in the clinical management of a small number of the women during subsequent admissions. Age, mean arterial pressure (MAP) and heart rate (HR) are known to be closely associated with AIX% in other groups of non-pregnant individuals and
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
similar associations has also been confirmed in pregnancy (13). We accounted for the influence of HR by using AIX% standardised to a heart rate of 75 beats per minute in our analysis, but our small sample size limited the number of other variables that could be adjusted for in the analysis. Hence, we employed the statistical technique of ‘‘forward selection’’ to select confounding factors for our adjusted analyses. Only maternal age, gestation and uric acid were selected for inclusion in the adjusted analysis of the primary outcome of timeto-delivery; although MAP was selected for inclusion in three of the five adjusted analyses relating to our secondary outcomes. We calculated MAP (1/3 systolic + 2/3 diastolic brachial blood pressure), rather than using the SphygmoCor derived value, so as to reflect the information potentially available from routine clinical assessment. The Omron M5-I/HEM-757 oscillometric device has been formally validated for general use, but it has not been specifically validated in pregnancy (20). The SphygmoCor generalised transfer function, used to generate the central aortic pulse wave, has not been validated against invasive intra-aortic manometry in pregnancy (6,7). This is unlikely to change due to ethical concerns about undertaking such an invasive procedure in pregnant women. Our study is also limited by its small sample size which means that our 95%CI’s are wide and we could not exclude the possibility that the differences in AIX%, related to severe preeclampsia and the induction of labour for preeclampsia, were not due to chance. This was a preliminary study intended to support a proposal for a larger study. While it was possible to identify hypertensive women at risk of developing complications in our antenatal triage ward setting, the admission criteria and variability in clinical practice may limit the generalisability to other hospital settings. Furthermore, the small number of women enrolled in the study means that our findings should be interpreted with caution. Although all of the women in the study were being assessed for hypertension in pregnancy, we categorised five women as being normotensive at initial study assessment. This is because brachial blood pressure measurements were based on the average of three study readings, which were sometimes lower than the ‘‘routine clinical measurements’’ recorded. This may have been because women had been resting for a prolonged period of time on the ward before study assessment. Only the initial study assessment involved a single individual measuring brachial BP with a single device, whereas the secondary outcomes relating to preeclampsia depend on brachial blood pressure measured as part of routine obstetric care. The assessment of proteinuria was based on urinary dipstick analysis, rather than 24-h urine collection. On labour ward, proteinuria was measured using an automated device (Siemens Clinitek Status Analyser, Medical Solutions Diagnostics, Tarrytown, NY), but elsewhere assessment of proteinuria was based on the visual assessment of urinary dipstick. We did not measure PWV in our study as substantial changes in PWV do not appear to be major component of the haemodynamic changes that occur in pregnancy (13). Comparison with Existing Literature Although no previous studies have assessed the relationship of AIX% with subsequent outcomes in women with pregnancies complicated by
485
486
G. Fullerton et al.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
hypertension, our findings are comparable with previous cross-sectional studies that have measured AIX% in women with gestational hypertension and preeclampsia. We found central AIX% to be 8.8% (95%CI 0.1–17.5) higher in women with preeclampsia compared to those with gestational hypertension. This is comparable, although somewhat lower, than three previous studies which found AIX% to be between 12% and 15% higher in preeclampsia compared with gestational hypertension (8,9,16). Our ROC AUC of 0.80 (95%CI 0.59–1.00) for the ability of AIX% to predict preeclampsia is also comparable with a previous case-control study which reported an AUC of 0.87 (95%CI 0.82–0.92) (25). Implications for Future Research Our preliminary study suggests that the additional measurement of AIX% in women being assessed for hypertension in pregnancy may be useful in differentiating between those women who can be safely discharged home from hospital and those who require more careful monitoring. In the United Kingdom, current clinical practice regarding management of women with hypertensive disorders of pregnancy is based on the National Institute for Health and Clinical Excellence (NICE) recommendations (26). Although detailed guidelines are available for the monitoring and treatment of overt gestational hypertension and preeclampsia, the literature is less clear regarding the management of borderline hypertension in pregnancy. AIX% opens up the possibility of outpatient or community-based screening to identify women at risk of developing complications with the potential to reduce hospital bed occupancy and the related stress and anxiety for the women concerned. A much larger cohort study is required to confirm our findings and to determine if AIX% is useful in guiding the subsequent obstetric monitoring and intervention in pregnant women with hypertension.
DECLARATION OF INTEREST The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
REFERENCES 1. Poppas A, Shroff SG, Korcarz CE, et al. Serial assessment of the cardiovascular system in normal pregnancy. Role of arterial compliance and pulsatile arterial load. Circulation 1997;20;95:2407–15. 2. Powe CE, Levine RJ, Karumanchi SA. Preeclampsia, a disease of the maternal endothelium: the role of antiangiogenic factors and implications for later cardiovascular disease. Circulation 2011;123:2856–69. 3. Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010;376:631–44. 4. Savvidou MD, Hingorani AD, Tsikas D, et al. Endothelial dysfunction and raised plasma concentrations of asymmetric dimethylarginine in pregnant women who subsequently develop pre-eclampsia. Lancet 2003;361:1511–17. 5. Nichols WW, O’Rourke, MF, Vlachopoulos, C. McDonald’s blood flow in arteries: theoretical, experimental and clinical principles. London: Hodder Arnold; 2011. 6. Gallagher D, Adji A, O’Rourke MF. Validation of the transfer function technique for generating central from peripheral upper limb pressure waveform. Am J Hypertens 2004;17:1059–67.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
7. Pauca AL, O’Rourke MF, Kon ND. Prospective evaluation of a method for estimating ascending aortic pressure from the radial artery pressure waveform. Hypertension 2001;38:932–7. 8. Avni B, Frenkel G, Shahar L, et al. Aortic stiffness in normal and hypertensive pregnancy. Blood Press 2010;19:11–15. 9. Elvan-Taspinar A, Franx A, Bots ML, et al. Central hemodynamics of hypertensive disorders in pregnancy. Am J Hypertens 2004;17:941–6. 10. Kaihura C, Savvidou MD, Anderson JM, et al. Maternal arterial stiffness in pregnancies affected by preeclampsia. Am J Physiol Heart Circ Physiol 2009;297: 759–64. 11. Khalil A, Jauniaux E, Cooper D, Harrington K. Pulse wave analysis in normal pregnancy: a prospective longitudinal study. PLoS One 2009;4:e6134. 12. Khalil AA, Cooper DJ, Harrington KF. Pulse wave analysis: a preliminary study of a novel technique for the prediction of pre-eclampsia. BJOG 2009;116:268–76. 13. Macedo ML, Luminoso D, Savvidou MD, et al. Maternal wave reflections and arterial stiffness in normal pregnancy as assessed by applanation tonometry. Hypertension 2008;51:1047–51. 14. Ronnback M, Lampinen K, Groop PH, Kaaja R. Pulse wave reflection in currently and previously preeclamptic women. Hypertens Pregn 2005;24:171–80. 15. Savvidou MD, Kaihura C, Anderson JM, Nicolaides KH. Maternal arterial stiffness in women who subsequently develop pre-eclampsia. PLoS One 2011;6:e18703. 16. Spasojevic M, Smith SA, Morris JM, Gallery ED. Peripheral arterial pulse wave analysis in women with pre-eclampsia and gestational hypertension. BJOG 2005; 112:1475–8. 17. Wilkinson IB, Hall IR, MacCallum H, et al. Pulse-wave analysis: clinical evaluation of a noninvasive, widely applicable method for assessing endothelial function. Arterioscler Thromb Vasc Biol 2002;22:147–52. 18. Soiza RL, Bhat MY, Crilly MA, Williams DJ. Nonequivalence of inhaled and nebulized salbutamol in the assessment of endothelial dysfunction. Cardiovasc Ther 2008;26:262–8. 19. Vlachopoulos C, Aznaouridis K, O’Rourke MF, et al. Prediction of cardiovascular events and all-cause mortality with central haemodynamics: a systematic review and meta-analysis. Eur Heart J 2010;31:1865–71. 20. El Assaad MA, Topouchian JA, Asmar RG. Evaluation of two devices for selfmeasurement of blood pressure according to the international protocol: the Omron M5-I and the Omron 705IT. Blood Press Monit 2003;8:127–33. 21. Crilly M, Coch C, Clark H, et al. Repeatability of the measurement of augmentation index in the clinical assessment of arterial stiffness using radial applanation tonometry. Scand J Clin Lab Invest 2007;67:413–22. 22. Brown MA, Lindheimer MD, de Swiet M, et al. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregn 2001;20: IX–XIV. 23. Wilkinson IB, MacCallum H, Flint L, et al. The influence of heart rate on augmentation index and central arterial pressure in humans. J Physiol 2000;525: 263–70. 24. Bonellie S, Chalmers J, Gray R, et al. Centile charts for birthweight for gestational age for Scottish singleton births. BMC Pregn Childbirth 2008;8:5. 25. Khalil A, Cowans NJ, Spencer K, et al. First-trimester markers for the prediction of pre-eclampsia in women with a-priori high risk. Ultrasound Obstet Gynecol 2010; 35:671–9. 26. National Collaborating Centre for Women’s and Children’s Health. Hypertension in pregnancy: the management of hypertensive disorders during pregnancy. London: National Institute for Health and Clinical Excellence (NICE) clinical guideline 107; 2011.
487
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
ORIGINAL ARTICLE
Measurement of aortic augmentation index in pregnant women with raised blood pressure and subsequent outcomes: a preliminary prospective cohort study Gail Fullerton,1 Michael A. Crilly,2 Sohinee Bhattacharya,2 and Peter J. Danielian1 1
Department of Obstetrics, Aberdeen Maternity Hospital, Foresterhill, Aberdeen, UK and 2 Division of Applied Health Sciences, University of Aberdeen, Foresterhill, Aberdeen, UK Objective: Preeclampsia is associated with arterial dysfunction and augmentation index (AIX%) is an established indicator of arterial dysfunction. Our aim was to investigate the relationship of AIX% with time-to-delivery and other outcomes in women admitted to an antenatal triage unit. Methods: We recruited 28 women with singleton pregnancies attending antenatal triage ward for assessment of hypertension. After 10 min rest, seated brachial blood pressure (Omron HEM-757) and AIX% (SphygmoCor applanation tonometry pulse wave analysis, PWA) were measured by a single investigator; other clinicians remained blinded to PWA results. Routine assessment included cardiotocography, urine analysis and blood tests. Subsequent outcomes were extracted from the obstetric records. Results: Mean AIX% was 19.7% (SD 11.5; range 4% to +36%), maternal age 31 years, gestation 37 weeks, brachial BP 145/95, proteinuria 39%. Nine women had preeclampsia at assessment and six subsequently developed preeclampsia. Median time-to-delivery was 10 d (IQR 1.6–25 d) and was shorter for AIX% 20% (median 8.9 versus 19.8 d). AIX% was higher with preeclampsia (24.0%; SD 9.5) versus gestational hypertension (15.2%; SD 12.4); absolute difference 8.8% (95%CI 0.1–17.5; p = 0.05). A one-point higher AIX% (adjusted for age, urate and gestation) was associated with 0.3 d (95%CI 0.5 to 0.0; p = 0.06) reduced time-to-delivery. A higher AIX% was also associated with induction for preeclampsia, severe preeclampsia, peripartum–antihypertensives and discharge-on-anti-hypertensives. Area under the curve (AUC) for AIX% predicting preeclampsia was 0.80 (95%CI 0.59–1.00; p = 0.04). Conclusion: AIX% is associated with time-to-delivery and other outcomes in pregnancy. Keywords Augmentation index, Applanation tonometry, Hypertension, Pregnancy, Preeclampsia, Pulse wave analysis.
Correspondence: Dr Michael A. Crilly, MD MPH MRCGP MFPHM, Senior Lecturer in Clinical Epidemiology, Division of Applied Health Sciences, University of Aberdeen, Polwarth Building, Foresterhill, Aberdeen AB25 2ZD, UK. Tel: +44(0)1224 437855. Fax: +44(0)1224 550925. E-mail: [email protected]
Augmentation index and preeclampsia
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
INTRODUCTION The major temporal haemodynamic changes that occur in normal pregnancy include an increase in arterial wall compliance, reduced arterial resistance and a reduction in the magnitude of arterial wave reflection within the arterial tree (1). This is in contrast with the widespread endothelial dysfunction and increased peripheral arterial resistance that is observed in pregnancies complicated by preeclampsia (2–4). Preeclampsia is an important indication for the induction of labour and a cause of premature delivery (2). Arterial wave reflection can be assessed non-invasively using pulse waveform analysis (PWA) combined with applanation tonometry at the radial artery. Using commercially available PWA devices, such as the SphygmoCor system, it is possible to rapidly assess central aortic pulse waves and pressures in routine clinical settings without the need for cardiac catheterisation (5–7). SphygmoCor PWA can be used to measure the augmentation index (AIX%), which depends upon the extent of arterial wave reflection (5). PWA is becoming an increasingly popular technique for assessing haemodynamics in pregnancy (8–16). The out-going systolic pulse wave generated by the heart in the conduit and peripheral arteries is also reflected back towards the heart where it can add to (‘‘augment’’) the central aortic pressure (5). The amount by which the central aortic pressure is increased by this phenomenon of wave reflection is the ‘‘augmentation pressure’’; the AIX% is the ‘‘augmentation pressure’’ (AP) expressed as a proportion of the central aortic ‘‘pulse pressure’’ (PP). AIX% represents a composite index of arterial function that is influenced by endothelial dysfunction, arterial compliance, peripheral arterial resistance and left ventricular ejection (5). Previous research in non-pregnant individuals has demonstrated that AIX% is closely associated with endothelial dysfunction (17,18). In non-pregnant populations, a higher AIX% has been shown to be associated with a higher rate of major adverse cardiovascular events such as myocardial infarction and stroke (19). Recent PWA studies in uncomplicated pregnancies have demonstrated that AIX% declines in the first half of pregnancy, reaches its nadir around midpregnancy and then rises again to early pregnancy levels at term (12,13). Compared to normotensive pregnancies AIX% is substantially higher in pregnancies complicated by hypertension (8–10,14,16). AIX% is also higher in preeclamptic pregnancies compared to those complicated by gestational hypertension (8,9,16). The majority of previous studies have been crosssectional, but a single prospective cohort study has demonstrated that a higher AIX%, measured at around 13 weeks gestation, is associated with an increased risk of subsequently developing preeclampsia (12). However, no previous studies have assessed the relationship of AIX% with subsequent outcomes in women with pregnancies complicated by hypertension. We wondered if the additional measurement of AIX% in women being assessed for hypertension in the second half of pregnancy might be helpful in identifying those women destined for a less benign clinical course. If this were the case then measuring AIX% might be useful in guiding the subsequent obstetric monitoring and intervention in pregnant women with hypertension. The aim of this preliminary prospective cohort study is to assess the relationship between obstetric outcomes and AIX% measured non-invasively
477
478
G. Fullerton et al.
in pregnant women admitted to a triage unit for hypertension in a routine healthcare setting.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
MATERIALS AND METHODS Women with a singleton pregnancy who attended the antenatal triage ward at Aberdeen Maternity Hospital (January to August 2012) with suspected hypertensive disorder were enrolled in this study. Women attended the antenatal triage ward for the assessment of high blood pressure and/or proteinuria following either: hospital antenatal clinic assessment, community midwife assessment, or after contacting the unit complaining of symptoms associated with preeclampsia (such as frontal headache, visual disturbance or epigastric pain). Routine clinical assessment of women admitted to the antenatal triage ward included aural temperature, brachial blood pressure, dipstick urine analysis (Siemens Multistix-8SG), venous blood sample (full blood count, urea, creatinine, liver function tests, uric acid) and cardiotocography. When proteinuria was present on dipstick testing, a midstream specimen of urine was sent for microscopy and culture. After resting seated for 10 min, all participants were assessed in the seated position by a single investigator (G. F.) on a single occasion. Right brachial BP was measured in triplicate using the Omron HEM-757 IntelliSense monitor with a standard cuff, or large sized cuff if required. The Omron M5-I /HEM-757 (Omron Healthcare Inc., Lake Forest, IL) is an automated oscillometric BP monitor, which has previously been independently validated against international criteria (20). The mean systolic and diastolic BP based on three readings was entered into the SphygmoCor device. Non-invasive PWA was also undertaken with the participant seated using the SphygmoCor device (SCOR v 8.0, AtCor Medical Pty Ltd., Sydney, Australia) with a hand-held high-fidelity tonometer (Millar Tonometer, Houston, TX) applanated (applied to flatten but not occlude) at the right radial artery. The SphygmoCor device captures waves over a 10-s period to produce an average pulse wave contour. At least two high quality PWA measurements were obtained from each woman. The quality of PWA was assessed visually on the laptop computer screen and numerically using the devices in-built percentage quality index score (QI%; device default setting based on: average pulse height, pulse height variation, diastolic variation, waveform shape deviation and maximum rate of pressure rise in the peripheral waveform). The SphygmoCor device calibrates the radial arterial waveform against the brachial blood pressure. The central aortic pulse wave is derived using a validated ‘‘generalised transfer function’’ which has previously been validated against simultaneous central aortic pressures obtained invasively by direct catheterisation (6,7). We have previously demonstrated the high level of repeatability of applanation tonometry PWA (21). All women underwent study assessment on the antenatal triage ward at Aberdeen Maternity Hospital between 10 a.m. and 10 p.m. Pulse wave velocity (PWV) was not measured. At initial study assessment information was also obtained concerning age, ethnicity, gestation, previous medical history, family history, current medication, current smoking status, number of previous deliveries 24 weeks and body mass index (BMI) recorded at the first antenatal booking visit. Laboratory results for full blood count, urea, creatinine, liver function and uric acid were
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
measured on admission to the antenatal triage ward. The subsequent treatment plan at ‘‘antenatal triage ward admission’’ was determined by medical staff not affiliated with the study who remained blinded to all of the PWA results throughout the study. The investigator (G. F.) was not involved in the clinical management of the women during the ‘‘triage ward admission’’, but she was involved in the clinical management of a small number of the women during subsequent admissions. Post-partum the obstetric case notes were reviewed by an obstetrician (G. F.) to determine subsequent use of peripartum (delivery ± 48 h) anti-hypertensive therapy, discharge home on anti-hypertensives, indications for induction of labour, mode of delivery, birth weight, major complications of pregnancy and the use of intravenous anti-hypertensives/ magnesium sulphate during labour. The highest brachial blood pressure and level of proteinuria during the peripartum period was recorded. Statistical Analysis The hypertensive status of women in the study was categorised according to International Society for the Study of Hypertension in Pregnancy (ISSHP) criteria as: normotensive if brachial BP5140/90 mmHg; gestational hypertension if BP 140/90 mmHg without proteinuria; and preeclamptic if BP 140/ 90 mmHg with proteinuria 1 + in the absence of a urinary tract infection (22). ‘‘Induction for preeclampsia’’ was defined as the use of vaginal prostaglandins, artificial rupture of membranes (ARM) and intravenous oxytocin in women who were not in labour with a brachial BP 140/90 and 1 + proteinuria on dipstick analysis. AIX% has a positive value if augmentation by the reflected wave occurs before the attainment of peak systolic pressure and a negative value if augmentation occurs after the attainment of peak systolic pressure (5). Since AIX% in an individual varies with heart rate it is commonly standardised to a heart rate of 75 beats-per-minute and this standardised AIX% (produced by the SphygmoCor device) was used in all analyses (23). Our pre-specified primary outcome was time-to-delivery (the SphygmoCor device automatically records that date and time of assessment and the maternity records the date and time of delivery). Secondary outcomes included peripartum (delivery ± 48 h) anti-hypertensive therapy, discharge home on anti-hypertensive therapy, induction for preeclampsia and the development of ‘‘severe’’ preeclampsia (defined as a brachial BP 160/ 110 mmHg, or BP 140/90 plus proteinuria 3+) (3). Continuous data approximating a normal distribution are presented as means (standard deviation) and skewed data are presented as medians (interquartile range, IQR). Mean arterial pressure (MAP, mmHg) was calculated as 1/3 systolic + 2/ 3 diastolic brachial blood pressure. Adjusted and unadjusted differences in mean AIX% were calculated and statistical precision was indicated using 95% confidence intervals (95%CI). Adjusted differences in mean AIX% were obtained using multiple linear regression (MLR) with forward selection of variables based on a probability for F of p = 0.08 for entry and p = 0.10 for removal. Statistical analysis was undertaken using IBM SPSS Statistics (version 20, IBM Corp., Armonk, NY). All p values are two-sided. All women provided written informed consent. The study adhered to the Helsinki Declaration and was approved by the North of Scotland Research Ethics Committee (14 November 2011; reference 11/NS/0039).
479
480
G. Fullerton et al.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
RESULTS A total of 31 women admitted to the antenatal triage ward for the assessment of hypertensive disorder in pregnancy were invited to participate in this study. A satisfactory PWA trace could not be obtained for one woman, one woman was excluded because of vasculitis and one woman declined to participate. The following analysis relates to the remaining 28 women. All were singleton pregnancies. No women were in a labour at the time of study assessment and all women with proteinuria ( 1+) had negative urinary microscopy and culture. AIX% ranged from 4% to +36% with a mean of 19.7 (SD 11.5) and a mean quality index (QI%) for PWA of 70% (SD 14). Women with an AIX% 20% had a higher brachial blood pressure, higher BMI at booking, higher urate level and were more likely to be nulliparous (Table 1). At initial study assessment, mean maternal age was 31 years, gestation 37 weeks (57% less than 38 weeks gestation), brachial BP 145/95 mmHg and 39% had proteinuria ( 1+) on urinary analysis (Table 1). The majority of women were nulliparous (75%) and Caucasian (93%). Only two women (7%) admitted to being current smokers and 24 (86%) reported never having smoked; two women were ex-smokers. Four women (14%) had a previous history of gestational hypertension and one woman had a prior history of essential hypertension. No women had a previous history of preeclampsia. Nine (32%) women reported a frontal headache at the time of assessment. One woman had multi-morbidity with rheumatoid arthritis, pernicious anaemia and hyperthyroidism (treated with sulphasalazine, prednisolone and vitamin B12). Table 1. Characteristics of women at initial study assessment. Augmentation index (AIX)
Mean maternal age (SD), years Mean gestation (SD), weeks Mean BMI at booking (SD), kg/m2 Mean brachial systolic BP, mmHg Mean brachial diastolic BP, mmHg Mean arterial pressure (SD), mmHg Nulliparous Brachial BP 140/90 mmHg Proteinuria ISSHP status at study assessment: Normotensive Gestational hypertension Preeclampsia Mean uric acid (SD), mmol/L Mean total bilirubin (SD), mmol/L Mean alanine-aminotransferase, IU/L Mean c-glutamyltransferase (SD), IU/L Mean creatinine (SD), mmol/L Mean platelets (SD), 109/L
AIX 20%
AIX520%
All women
(n = 17)
(n = 11)
(N = 28)
30 37.0 29 146 96 115 15 16 7
(5) (3.3) (5) (16) (8) (10) 88% 94% 41%
32 36.7 27 142 90 106 6 7 4
(4) (3.4) (5) (16) (11) (12) 55% 64% 36%
31 36.9 28 145 94 111 21 23 11
(4) (3.3) (5) (16) (10) (11) 75% 82% 39%
1 9 7 345 6 11 14 53 262
6% 53% 41% (44) (2) (4) (12) (5) (47)
4 5 2 285 7 14 10 50 237
36% 46% 18% (48) (2) (4) (3) (11) (53)
5 14 9 321 7 12 13 52 252
18% 50% 32% (53) (2) (4) (10) (8) (50)
All values are numbers (percentages) unless otherwise indicated. SD, standard deviation; BMI, body mass index; BP, blood pressure; ISSHP, International Society for the Study of Hypertension in Pregnancy.
Augmentation index and preeclampsia
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Obstetric Outcomes Labour was induced for 21 (75%) women among whom 10 were induced for preeclampsia. Twelve women (46%) had a spontaneous vaginal delivery, 10 (38%) caesarean section and 6 (23%) instrumental delivery. One woman had premature rupture of membranes at 28 weeks and was delivered by caesarean section; one woman had a major post-partum haemorrhage; and one woman had a caesarean section for foetal bradycardia during labour. Two women delivered babies with a birth weight below the 3rd Scottish centile (standardised for gestation and sex) (24). No other major complication of pregnancy occurred and no women had seizures. Time-to-Delivery The median time-to-delivery after study assessment was 10 d (IQR 1.6–25 d) and ranged from 0.5 to 61 d. Women with an AIX% 20% had a shorter timeto-delivery (median 8.9 d) compared to women with an AIX%520% (median 19.8 d). Unadjusted and adjusted regression coefficients in relation to time-todelivery are shown in Table 2. On unadjusted analysis a one-point higher AIX% at assessment was associated with 0.5 d (95%CI 1.1 to 0.1; p = 0.08) reduction in the time-to-delivery. Forward variable selection identified AIX%, gestation, urate and maternal age at initial assessment for inclusion in the regression model, which explained 82% of the variability in the time-todelivery (Table 2). On adjusted analysis a one-point higher AIX% at assessment was associated with 0.3 d (95%CI 0.5 to 0.0; p = 0.06) reduction in the time-to-delivery. Gestation at study assessment was the most important determinant of time-to-delivery with each additional week of gestation being associated with an adjusted 4.4 d (95%CI 3.5–5.3; p50.001) reduction in the time-to-delivery. Hypertensive Status At initial assessment, based on ‘‘study’’ brachial BP measurement and ISSHP criteria, nine (32%) women had preeclampsia, 14 (50%) women gestational hypertension and 5 (18%) were normotensive (Table 3). Four (14%) women were taking anti-hypertensive therapy at the time of study assessment and a further 8 (29%) women were commenced on anti-hypertensive therapy shortly after PWA-assessment. Sixteen women (57%) received peripartum antihypertensive therapy (with either labetalol or methyldopa); two of these Table 2. Factors associated with the primary outcome of time-to-delivery in days. Unadjusted coefficient Augmentation index, AIX% Gestation, weeks Uric acid, mmol/L Maternal age, years
0.50 4.50 0.13 0.64
Adjusted coefficienta 0.25 4.37 0.07 0.59
95%CI
95%CI
0.52 5.29 0.13 1.24
0.01 3.45 0.01 0.06
p 0.06 550.001 0.03 0.07
Summary of regression model: R 0.92, adjusted R2 0.82, p (ANOVA)550.001. Variables excluded by SPSS ‘‘forward selection’’: mean arterial pressure, proteinuria, nulliparous, BMI at booking, total bilirubin, alanine-aminotransferase, gamma-glutamyltransferase, creatinine, platelets.
a
481
482
G. Fullerton et al. Table 3. Hypertensive status at initial assessment and peripartum. Peripartum ISSHP status:
ISSHP status at assessment:
Normotensive
Gestational
Preeclampsia
Total nos (%)
Normotensive Gestational Preeclampsia Total nos (%)
1 1 0 2 (7%)
2 9 3 14 (50%)
2 4 6 12 (43%)
5 (18) 14 (50) 9 (32) 28 (100%)
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
ISSHP, International Society for the Study of Hypertension in Pregnancy.
women required intra-venous anti-hypertensive therapy during labour and one of these two women also received magnesium sulphate. Eleven (39%) women were discharged home on oral anti-hypertensive therapy. Change in hypertensive status between study assessment and the peripartum period (delivery ± 48 h) is shown in Table 3; the shaded boxes indicate the number of women whose hypertensive status remained unchanged. Hypertensive status remained unchanged for 16 (57%) women, worsened for 8 (29%) and improved for 4 (14%) women. Only one woman (study BP 135/89 with proteinuria and an elevated uric acid of 360 mmol/L) remained normotensive throughout pregnancy. Overall, 15 women met ISSHP criteria for preeclampsia during pregnancy, including 10 with ‘‘severe’’ preeclampsia (five at initial study assessment and five at peripartum); the remaining 12 women had gestational hypertension. AIX% was 24.0% (SD 9.5) in women with preeclampsia and 15.2% (SD 12.4) in women with gestational hypertension; an absolute difference of 8.8% (95%CI 0.1–17.5; p = 0.05). AIX% was highest among the five women with ‘‘severe’’ preeclampsia at initial study assessment (26.4%, SD 5.4). Secondary Outcomes The relationship of AIX% at initial study assessment with secondary outcomes is shown in Table 4. On adjusted analysis, AIX% was an absolute 10% higher in women with preeclampsia, those who require peripartum anti-hypertensive therapy and discharged home on anti-hypertensive therapy. Although not statistically significant at the 5% level, AIX% was also 8% higher among women induced for preeclampsia and among those with ‘‘severe’’ preeclampsia. Nine women met ISSHP criteria for preeclampsia at initial study assessment. Among the remaining 19 women free of preeclampsia at study assessment, 6 (32%) subsequently developed preeclampsia. Figure 1 shows the receiver operating characteristic (ROC) curve for these 19 women. The area under the curve (AUC) for the ability of AIX% 20% (i.e. above the mean AIX% value for the study) to predict the subsequent occurrence of preeclampsia was 0.80 (95%CI 0.59–1.00; p = 0.04).
DISCUSSION Among the women in this study being assessed for hypertension in pregnancy, a higher AIX% was independently associated with a shorter time-to-delivery after adjusting for maternal age, urate and gestation at baseline assessment
24.1 24.3 24.0 23.5 25.3
(10.9) (10.5) (9.5) (9.0) (9.2)
13.8 16.2 14.7 17.6 16.6
(10.0) (11.4) (12.0) (12.5) (11.7)
Mean AIX% (SD) Mean AIX% (SD)
Absent
10.3 8.1 9.3 5.9 8.8
Crude 10.4 9.6 9.9 8.0 7.5 b
b
b
a
a
Adjusted
Mean difference AIX
(2.4–18.2) (1.4–17.8) (2.1–17.6) ( 0.7 to 16.7) ( 1.3 to 16.2)
(95%CI)
0.01 0.02 0.02 0.07 0.09
p
0.55 0.52 0.56 0.49 0.47
R
0.24 0.21 0.31 0.18 0.16
Adj R2
0.01 0.02 0.01 0.03 0.04
p (ANOVA)
Summary regression model
Variables included by SPSS ‘‘forward selection’’: nulliparitya and mean arterial blood pressure b at initial study assessment. [Excluded variables (measured at initial study assessment): maternal age, gestation, proteinuria, uric acid, total bilirubin, alanine-aminotransferase, gammaglutamyltransferase, creatinine, platelets, body mass index at booking].
Peripartum anti-hypertensive therapy (16/12) Home on anti-hypertensive therapy (12/16) Preeclampsia (15/13) Severe preeclampsia (10/18) Induced for preeclampsia (10/18)
Secondary outcomes (present/absent)
Present
Table 4. Augmentation index (AIX%) at initial assessment and secondary outcomes.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
483
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
484
G. Fullerton et al.
Figure 1. Receiver operating characteristic (ROC) curve for augmentation index (AIX% 20%) and the occurrence of preeclampsia among women initially free of preeclampsia (n = 19).
(one-point increase in AIX% was associated with a 0.3 d [95%CI 0.5 to 0.0; p = 0.06] reduction in the time-to-delivery). AIX% was higher in women with preeclampsia compared to gestational hypertension. AIX% was also predictive of the subsequent development of preeclampsia after assessment. A higher AIX% was associated with induction of labour for preeclampsia, severe preeclampsia, peripartum-anti-hypertensive therapy and being discharged home on anti-hypertensive therapy. Study Strengths and Limitations Our study has some important strengths as it was undertaken by a single investigator in a routine healthcare setting using equipment that has previously been validated for the measurement of brachial BP (Omron M5-I/ HEM-757) (20) and PWA (SphygmoCor) (6,7). In order to ensure that women were haemodynamically stable at the time of assessment, PWA was undertaken after women had rested for at least 10 min. So as to avoid the knowledge of AIX% levels influencing clinical management, the medical staff providing obstetric care remained unaware of the PWA measurements, although the obstetrician (G. F.) undertaking PWA was involved in the clinical management of a small number of the women during subsequent admissions. Age, mean arterial pressure (MAP) and heart rate (HR) are known to be closely associated with AIX% in other groups of non-pregnant individuals and
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
similar associations has also been confirmed in pregnancy (13). We accounted for the influence of HR by using AIX% standardised to a heart rate of 75 beats per minute in our analysis, but our small sample size limited the number of other variables that could be adjusted for in the analysis. Hence, we employed the statistical technique of ‘‘forward selection’’ to select confounding factors for our adjusted analyses. Only maternal age, gestation and uric acid were selected for inclusion in the adjusted analysis of the primary outcome of timeto-delivery; although MAP was selected for inclusion in three of the five adjusted analyses relating to our secondary outcomes. We calculated MAP (1/3 systolic + 2/3 diastolic brachial blood pressure), rather than using the SphygmoCor derived value, so as to reflect the information potentially available from routine clinical assessment. The Omron M5-I/HEM-757 oscillometric device has been formally validated for general use, but it has not been specifically validated in pregnancy (20). The SphygmoCor generalised transfer function, used to generate the central aortic pulse wave, has not been validated against invasive intra-aortic manometry in pregnancy (6,7). This is unlikely to change due to ethical concerns about undertaking such an invasive procedure in pregnant women. Our study is also limited by its small sample size which means that our 95%CI’s are wide and we could not exclude the possibility that the differences in AIX%, related to severe preeclampsia and the induction of labour for preeclampsia, were not due to chance. This was a preliminary study intended to support a proposal for a larger study. While it was possible to identify hypertensive women at risk of developing complications in our antenatal triage ward setting, the admission criteria and variability in clinical practice may limit the generalisability to other hospital settings. Furthermore, the small number of women enrolled in the study means that our findings should be interpreted with caution. Although all of the women in the study were being assessed for hypertension in pregnancy, we categorised five women as being normotensive at initial study assessment. This is because brachial blood pressure measurements were based on the average of three study readings, which were sometimes lower than the ‘‘routine clinical measurements’’ recorded. This may have been because women had been resting for a prolonged period of time on the ward before study assessment. Only the initial study assessment involved a single individual measuring brachial BP with a single device, whereas the secondary outcomes relating to preeclampsia depend on brachial blood pressure measured as part of routine obstetric care. The assessment of proteinuria was based on urinary dipstick analysis, rather than 24-h urine collection. On labour ward, proteinuria was measured using an automated device (Siemens Clinitek Status Analyser, Medical Solutions Diagnostics, Tarrytown, NY), but elsewhere assessment of proteinuria was based on the visual assessment of urinary dipstick. We did not measure PWV in our study as substantial changes in PWV do not appear to be major component of the haemodynamic changes that occur in pregnancy (13). Comparison with Existing Literature Although no previous studies have assessed the relationship of AIX% with subsequent outcomes in women with pregnancies complicated by
485
486
G. Fullerton et al.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
hypertension, our findings are comparable with previous cross-sectional studies that have measured AIX% in women with gestational hypertension and preeclampsia. We found central AIX% to be 8.8% (95%CI 0.1–17.5) higher in women with preeclampsia compared to those with gestational hypertension. This is comparable, although somewhat lower, than three previous studies which found AIX% to be between 12% and 15% higher in preeclampsia compared with gestational hypertension (8,9,16). Our ROC AUC of 0.80 (95%CI 0.59–1.00) for the ability of AIX% to predict preeclampsia is also comparable with a previous case-control study which reported an AUC of 0.87 (95%CI 0.82–0.92) (25). Implications for Future Research Our preliminary study suggests that the additional measurement of AIX% in women being assessed for hypertension in pregnancy may be useful in differentiating between those women who can be safely discharged home from hospital and those who require more careful monitoring. In the United Kingdom, current clinical practice regarding management of women with hypertensive disorders of pregnancy is based on the National Institute for Health and Clinical Excellence (NICE) recommendations (26). Although detailed guidelines are available for the monitoring and treatment of overt gestational hypertension and preeclampsia, the literature is less clear regarding the management of borderline hypertension in pregnancy. AIX% opens up the possibility of outpatient or community-based screening to identify women at risk of developing complications with the potential to reduce hospital bed occupancy and the related stress and anxiety for the women concerned. A much larger cohort study is required to confirm our findings and to determine if AIX% is useful in guiding the subsequent obstetric monitoring and intervention in pregnant women with hypertension.
DECLARATION OF INTEREST The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
REFERENCES 1. Poppas A, Shroff SG, Korcarz CE, et al. Serial assessment of the cardiovascular system in normal pregnancy. Role of arterial compliance and pulsatile arterial load. Circulation 1997;20;95:2407–15. 2. Powe CE, Levine RJ, Karumanchi SA. Preeclampsia, a disease of the maternal endothelium: the role of antiangiogenic factors and implications for later cardiovascular disease. Circulation 2011;123:2856–69. 3. Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010;376:631–44. 4. Savvidou MD, Hingorani AD, Tsikas D, et al. Endothelial dysfunction and raised plasma concentrations of asymmetric dimethylarginine in pregnant women who subsequently develop pre-eclampsia. Lancet 2003;361:1511–17. 5. Nichols WW, O’Rourke, MF, Vlachopoulos, C. McDonald’s blood flow in arteries: theoretical, experimental and clinical principles. London: Hodder Arnold; 2011. 6. Gallagher D, Adji A, O’Rourke MF. Validation of the transfer function technique for generating central from peripheral upper limb pressure waveform. Am J Hypertens 2004;17:1059–67.
Hypertens Pregnancy Downloaded from informahealthcare.com by University of Bristol on 02/21/15 For personal use only.
Augmentation index and preeclampsia
7. Pauca AL, O’Rourke MF, Kon ND. Prospective evaluation of a method for estimating ascending aortic pressure from the radial artery pressure waveform. Hypertension 2001;38:932–7. 8. Avni B, Frenkel G, Shahar L, et al. Aortic stiffness in normal and hypertensive pregnancy. Blood Press 2010;19:11–15. 9. Elvan-Taspinar A, Franx A, Bots ML, et al. Central hemodynamics of hypertensive disorders in pregnancy. Am J Hypertens 2004;17:941–6. 10. Kaihura C, Savvidou MD, Anderson JM, et al. Maternal arterial stiffness in pregnancies affected by preeclampsia. Am J Physiol Heart Circ Physiol 2009;297: 759–64. 11. Khalil A, Jauniaux E, Cooper D, Harrington K. Pulse wave analysis in normal pregnancy: a prospective longitudinal study. PLoS One 2009;4:e6134. 12. Khalil AA, Cooper DJ, Harrington KF. Pulse wave analysis: a preliminary study of a novel technique for the prediction of pre-eclampsia. BJOG 2009;116:268–76. 13. Macedo ML, Luminoso D, Savvidou MD, et al. Maternal wave reflections and arterial stiffness in normal pregnancy as assessed by applanation tonometry. Hypertension 2008;51:1047–51. 14. Ronnback M, Lampinen K, Groop PH, Kaaja R. Pulse wave reflection in currently and previously preeclamptic women. Hypertens Pregn 2005;24:171–80. 15. Savvidou MD, Kaihura C, Anderson JM, Nicolaides KH. Maternal arterial stiffness in women who subsequently develop pre-eclampsia. PLoS One 2011;6:e18703. 16. Spasojevic M, Smith SA, Morris JM, Gallery ED. Peripheral arterial pulse wave analysis in women with pre-eclampsia and gestational hypertension. BJOG 2005; 112:1475–8. 17. Wilkinson IB, Hall IR, MacCallum H, et al. Pulse-wave analysis: clinical evaluation of a noninvasive, widely applicable method for assessing endothelial function. Arterioscler Thromb Vasc Biol 2002;22:147–52. 18. Soiza RL, Bhat MY, Crilly MA, Williams DJ. Nonequivalence of inhaled and nebulized salbutamol in the assessment of endothelial dysfunction. Cardiovasc Ther 2008;26:262–8. 19. Vlachopoulos C, Aznaouridis K, O’Rourke MF, et al. Prediction of cardiovascular events and all-cause mortality with central haemodynamics: a systematic review and meta-analysis. Eur Heart J 2010;31:1865–71. 20. El Assaad MA, Topouchian JA, Asmar RG. Evaluation of two devices for selfmeasurement of blood pressure according to the international protocol: the Omron M5-I and the Omron 705IT. Blood Press Monit 2003;8:127–33. 21. Crilly M, Coch C, Clark H, et al. Repeatability of the measurement of augmentation index in the clinical assessment of arterial stiffness using radial applanation tonometry. Scand J Clin Lab Invest 2007;67:413–22. 22. Brown MA, Lindheimer MD, de Swiet M, et al. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregn 2001;20: IX–XIV. 23. Wilkinson IB, MacCallum H, Flint L, et al. The influence of heart rate on augmentation index and central arterial pressure in humans. J Physiol 2000;525: 263–70. 24. Bonellie S, Chalmers J, Gray R, et al. Centile charts for birthweight for gestational age for Scottish singleton births. BMC Pregn Childbirth 2008;8:5. 25. Khalil A, Cowans NJ, Spencer K, et al. First-trimester markers for the prediction of pre-eclampsia in women with a-priori high risk. Ultrasound Obstet Gynecol 2010; 35:671–9. 26. National Collaborating Centre for Women’s and Children’s Health. Hypertension in pregnancy: the management of hypertensive disorders during pregnancy. London: National Institute for Health and Clinical Excellence (NICE) clinical guideline 107; 2011.
487
