Hypertension

A 3-year-old attends the ED with a vomiting illness.  Routine observations include a blood pressure measured consistently as 125/80.

What is the definition of hypertension?

Blood pressure is notoriously difficult to measure in children. The definition of hypertension is a fluid concept but generally accepted to be >95^th or >99^th centile on standardised charts (established with data from the 3rd Taskforce on Childhood Hypertension, now in its 4th incarnation) corrected for age, sex, and height.

How do we measure blood pressure?

The ‘textbook’ measurement of BP is done after the patient has rested for 3 minutes, while sitting, and the cuff and manometer are at the level of the heart. The cuff should be an appropriate size – sufficient to cover 2/3 of the upper arm. The inflation bladder should cover 80-100% (preferably 100%) of the arm circumference.

Too small a cuff gives an artificially HIGH reading.

Too large a cuff gives an artificially LOW reading.

Measurement by automated machines using computed algorithms is easy, and does not require much training, but is poorly reproducible and unreliable. Manual BP measurement is superior, particularly if combined with a ‘green light’ system, or an audible trigger with a portable Doppler. ‘Gold standard’ is invasive measurement with an in-dwelling arterial line. This is clearly not appropriate for routine assessment.

Some centres advocate the use of ambulatory BP recordings, providing a 24-hour snapshot of BP readings, less likely to be influenced by anxiety, and able to demonstrate any consistent elevation in BP. ABPM is limited to older children, suffers from a similar level of inaccuracy to all machine-derived readings, and requires interpretation by someone familiar with them.

Taking all this into account, it can be seen that accurate measurement of BP in a younger child is a difficult task, and almost impossible in a single visit, particularly if the child is unwell.

Assuming that BP has been measured correctly and is high, is it physiological or pathological?

Physiological rise in BP – anxiety and pain are the commonest reasons for the elevation of BP in the ED. This is why measurement in a relaxed state becomes essential. There may also be a rise in BP to compensate for vascular underfilling when dehydrated – a fall in BP is a much later finding.

Pathological – repeated measurements (including on ABPM if available) will be elevated, regardless of analgesic control or emotional state. Persistence of hypertension when anaesthetised virtually guarantees pathological hypertension, as most anaesthetic agents vasodilate and have an anti-hypertensive effect.

Acute presentation – some children may present acutely with signs of severe hypertension – headache is classical. Seizures are not uncommon. Presentation with a facial nerve palsy demands measurement of BP. Undiagnosed severe hypertension may present rarely with signs of stroke. Where symptomatic, lowering of BP to an asymptomatic range is an emergency requiring IV medication and at least a high-dependency level of care but beware of lowering the BP too rapidly.

Aim to reduce presentation BP by 1/3 in the first 24 hours, 1/3 in the next 24 hours, and the final 1/3 over 48.

So: presenting BP of 160mmHg in a 6-year-old with headaches, target BP 100mmHg

• Lower BP to 140 on day 1
• Lower BP to 120 on day 2
• Optimal BP of 100 by day 4

What further investigations should I do?

Additional tests have two purposes – to identify the underlying cause, and to look for secondary effects.

Investigations for Cause

Renal ultrasound – may demonstrate scarring of kidneys related to previous infection, vesico-ureteric reflux, or dysplasia. Any geographic abnormality e.g. horseshoe kidney, cross-fused ectopia, may be associated with hypertension.

Doppler ultrasound – all imaging of the kidneys should include a Doppler assessment of the renal blood supply. This may identify renal artery stenosis (RAS), one of the commoner causes of hypertension.

Magnetic resonance arteriography – where there is a high suspicion that hypertension is related to RAS but Dopplers are normal, MRA may show stenosis of the arterial supply that cannot be visualised on ultrasound, especially intra-renal stenosis.

Serum renin and aldosterone – elevation can confirm renal production driving hypertension. If indicated, localisation can be assisted by performing intra-operative renal vein sampling.  This is restricted to specialist centres and only indicated in severe hypertension.

Urinary/plasma catecholamines – phaeochromocytoma not infrequently presents with incidental hypertension. Elevation of catecholamines and methylated derivatives in urine and plasma is highly suggestive of such a tumour, which may not be identified on initial ultrasound.

Glomerulonephritis investigation – if the elevation in BP is associated with haematuria, proteinuria, or renal impairment, investigations should include complement screening (C3, C4), anti-streptolysin O titre and anti-DNase B, auto-antibodies (including anti-GBM and ANCA). Nephrology referral should be considered early where hypertension is associated with evidence of acute ongoing kidney disease.

Investigations for consequence

Echocardiogram for left ventricular hypertension

Ophthalmic assessment for retinal changes

Urinalysis for proteinuria

These indicate more chronic hypertension and can provide further proof that a raised BP is not an artefactual finding. Their absence does NOT exclude pathological hypertension.

What’s the best management?

In a child with moderately elevated BP, management can be oral anti-hypertensives and carried out with an out-patient review. If BP is significantly elevated or symptomatic, consider admission and CAUTIOUS treatment of BP to a more normal range over several days using an intravenous agent. IV anti-hypertensives can be carefully adjusted and over-treatment avoided, whilst oral medicines tend to be longer-acting. A rapid drop in BP cannot easily be reversed.

The aim of treatment should be to restore BP to within normal centiles. Many authorities suggest that aiming for <95th centile is acceptable, while others that the 50th centile should be the target. There is universal agreement that BP management should be more aggressive in the diabetic population, where the 50th-75th centile is the maximum BP that should be permitted.

Choice of Agent – IV

Labetalol is the commonest agent used IV to correct severe hypertension. It has the advantage of clinician familiarity and is relatively cheap.

Sodium nitroprusside is a highly effective IV anti-hypertensive but is limited by conversion to cyanide restricting it to short-term use only.

Choice of agent – oral

General guidance is now that there are 4 classes of anti-hypertensives – ABCD drugs

A – ACE inhibitors are the logical choice where hypertension is related to the production of renin/aldosterone from a scarred kidney, or one trying to overcome arterial stenosis. Monitoring of renal function and hyperkalaemia is recommended upon starting treatment, and with dose increases.

B – Beta-blockers are the most established and familiar anti-hypertensive, with a long history of safety in children. Atenolol is often the oral drug of choice, due to its relatively long duration of action.

C – Calcium-channel blockers are used in two circumstances. Nifedipine is frequently used as a ‘stat’ medication to rapidly lower an elevated BP. This should be done cautiously if the duration of hypertension is long, or unknown. A precipitous drop in BP should be anticipated and access available to give 0.9% saline as volume to restore BP if this occurs.  Amlodipine is the commonest calcium-channel blocker used. Neonates and infants have a relatively rapid metabolism of amlodipine, so twice-daily dosing may be required.

D – Diuretics, such as furosemide, are useful where hypertension is driven by fluid overload.

Hydralazine may sometimes be added where adequate control of BP is not achieved for two or three of the above agents (and may be used IV in acute hypertension to stabilise). The side-effect profile of hydralazine makes it unsuitable for first-line use.

What’s the prognosis?

The prognosis depends on the cause. Essential hypertension (increasingly common in the increasingly large population) has a relatively poor prognosis given the intractable nature of obesity and poor response to dietary modification.

Prognosis in other conditions is dictated by how effectively BP can be controlled. Where the cause can be treated (e.g. renal artery stenosis amenable to stenting, resectable phaeochromocytoma) then the prognosis is excellent. Retinal and cardiac sequelae typically resolve over time when BP is normalised. Where BP remains elevated or requires multiple agents, the prognosis is more guarded, but morbidity tends to be an issue of adult life, and as such is poorly appreciated by paediatricians! 

So, should I ignore the high blood pressure of the 3-year-old?

NO!  Though high readings are often attributed to distress or crying, particularly in younger children, measurements should be repeated, ideally when the child is more relaxed (or even asleep).

Consistently high BP readings above the 90th centile merit some thought, and consistently above the 95th centile needs investigation.

References

Jorg R, Milani GP, Simonetti GD, Bianchetti MG, ‘Peripheral facial nerve palsy in severe systemic hypertension: a systematic review’ Am J Hypertens 2013;26: 351-356.

National High Blood Pressure Education Program Working Group on high blood pressure in children and adolescents ‘The Fourth Report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents’ Pediatrics 2004; 114: 555-576 https://www.nhlbi.nih.gov/health/prof/heart/hbp/hbp_ped.pdf

Soergel M, Kirschstein M, Busch C et al ‘Oscillometric twenty-four hour ambulatory blood pressure values in healthy children and adolescents: a multicenter trial involving 1141 subjects’ J Pediatr 1997; 130; 178-184.

Meyers K, Falkner B ‘Hypertension in children and adolescents: an approach to management of complex hypertension in pediatric patients’ Curr Hypertens Rep 2009; 11: 315-322.