Postgrad MedJ 1998;74:208-215 C The Fellowship of Postgraduate Medicine, 1998

New techniques

Assessment of respiratory musclefunction and strength

Nightingale Syabbalo

SummaryMeasurement of respiratory mus-cle strength is useful in order todetect respiratory muscle weak-ness and to quantify its severity. Inpatients with severe respiratorymuscle weakness, vital capacity isreduced but is a non-specific andrelatively insensitive measure.Conventionally, inspiratory andexpiratory muscle strength hasbeen assessed by maximal in-spiratory and expiratory mouthpressures sustained for 1 s (PImaxand PEmax) during maximalstatic manoeuvre against a closedshutter. However, Plmax andPEmax are volitional tests, andare poorly reproducible with anaverage coefficient of variation of25%. The sniff manoeuvre is natu-ral and probably easier to per-form. Sniff pressure, and snifftransdiaphragmatic pressure aremore reproducible and usefulmeasure of diaphragmaticstrength. Nevertheless, the sniffmanoeuvre is also volition-dependent, and submaximal ef-forts are most likely to occur inpatients who are ill or breathless.Non-volitional tests include meas-urements of twitch oesophageal,gastric and transdiaphragmaticpressure during bilateral electri-cal and magnetic phrenic nervestimulation. Electrical phrenicnerve stimulation is technicallydifficult and is also uncomfortableand painful. Magnetic phrenicnerve stimulation is less painfuland transdiaphragmatic pressureis reproducible in normal sub-jects. It is a relatively easy test thathas the potential to become awidely adopted method for theassessment of diaphragmstrength. The development of atechnique to measure diaphrag-matic sound (phonomyogram)during magnetic phrenic nervestimulation opens the way fornoninvasive assessment of dia-phragmatic function.

Keywords: phrenic nerve stimulation; respira-tory muscle strength; diaphragmatic fiuctionFaculty of Medicine, Kuwait University,PO Box 24923, 13110 Safat, KuwaitN Syabbalo

Correspondence to Dr N Syabbalo, Department ofClinical Physiology, College of Medicine, SultanQaboos University, PO Box 35, Al-Khod 123,Muscat, Oman

Accepted 23 September 1997

Respiratory muscle weakness is present in many patients and may represent asignificant problem for several reasons.' It is a common feature of severalneuromuscular diseases (box 1) and is associated with a higher morbidity andmortality.`5 Recently, there has been increasing awareness that respiratory mus-cle weakness can be a compounding factor in many diseases, such asmalnutrition,6 chronic obstructive pulmonary disease,` congestive heartfailure,'0-"2 and sepsis.'3 Furthermore, respiratory muscle weakness is a compli-cation of many metabolic diseases (box 2),'4 endocrine disorders, and steroidtherapy.'5 Measurement of respiratory muscle strength is useful in order todetect respiratory muscle weakness and to quantify its severity,'6 and todetermine optimal long-term management. Respiratory muscle testing is usefulin establishing the causes of unexplained dyspnoea, bulbar problems, andimpaired cough, and is also important for the follow-up of patients with respira-tory muscle weakness or those on respiratory muscle training programmes, andit is currently being used in ventilator weaning protocols."' Indications for evalu-ation of respiratory muscle strength are shown in box 3.

History and physical examination

Mild weakness of respiratory muscles or hemidiaphragmatic paralysis may beasymptomatic,' 1 or may present with dyspnoea in the presence of coexistingrespiratory disease. Isolated diaphragm weakness and even paralysis is withoutlife-threatening sequelae," which underlines the fact that, for ventilatory failureto occur, load must exceed capacity.20

Global, severe respiratory muscle weakness causes dyspnoea, initially duringexertion and eventually at rest.2' 22 Severe isolated weakness of the diaphragm orbilateral diaphragm paralysis causes orthopnoea when the patient is in the supineposition. Thus, patients with bilateral weakness of the diaphragm are unable tolie flat. Severe orthopnoea and obvious abdominal paradox occur only when thestrength of the diaphragm is reduced to approximately 25% of normal.2'

Severe respiratory muscle weakness may of itself cause ventilatory failure (eg,amyotrophic lateral sclerosis),23 but most commonly weakness is one of severalfactors contributing to ventilatory failure (eg, patients in an intensive care unitwith weaning difficulties). Ventilatory failure due to respiratory muscle weaknessusually develops gradually and hypoventilation first develops during sleep.' 24Such patients present with characteristic problems of disordered sleep, includingdaytime tiredness, somnolence, headache, and cognitive and intellectual impair-ment. Eventually ventilatory failure, pulmonary hypertension and cor pulmonaledevelop.25Some patients with neuromuscular diseases (box 1) have bulbar dysfunction

which predisposes to aspiration. Abdominal muscles are also weak and thisimpairs cough and clearance of secretions. These factors predispose to recurrentrespiratory infections. If neuromuscular disease spares the diaphragm,ventilatory failure is unlikely to occur. However, in generalised neuromusculardisorders it is unusual for the respiratory muscles to be spared.26

History and examination may confirm features of the many conditions, bothneurogenic and myopathic, that can cause respiratory muscle weakness.26 Wast-ing and fasciculation of the limb and accessory muscles can be striking, such as,for example, in motor neurone disease. An important specific symptom of severediaphragm weakness is dyspnoea when sitting or supine, or when standing inwater.The characteristic finding of profound bilateral diaphragm weakness or

paralysis is paradoxical movements of the rib cage and abdomen (abdominalparadox), and respiratory alternans, which are more obvious with the patientsupine.'9 There is active contraction of accessory muscles when ventilation isincreased. In the upright position, the abdominal muscles may be visiblyrecruited during expiration, serving to elevate the diaphragm and allowing

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from

Assessment of respiratory muscle function 209

Neuromuscular diseasesassociated with respiratorymuscles

Neurogenic* tetraplegia and paraplegia (trauma)* hemiplegia (stroke)* motor neurone disease* polyneuropathy* Guillain-Barre syndrome* Werdnig-HofEnan syndrome* amyotrophic lateral sclerosis* poliomyelitis* phrenic nerve paralysis (surgical

trauma, tumour infiltration)* 'frostbitten phrenic' (cardiac surgery,

cardioplegia)* transverse myelitis* multiple sclerosis* neuralgic amyotrophy* Parkinson's disease* Charcot-Marie-Tooth disease

Muscular* muscular atrophy* muscular dystrophy (Duchenne)* myotonic dystrophy (dystrophia

myotonica)* chronic obstructive pulmonary disease* congestive heart failure* inflammatory myopathies* systemic lupus erythematosus* rheumatoid arthritis* dermatomyositis* polymyositis* kyphoscoliosis

Neuromuscular junction* myasthenia gravis* Eaton-Lambert myasthenic syndrome* cholinergic crisis* botulinism* tetanus* sepsis* snake envenomation* tick paralysis

Drug-induced* aminoglycosides* barbiturates* anaesthetics* chloroquine* quinidine* dilantin* tricyclic antidepressants

Poisoning* organophosphates* lead* arsenic

Box 1

subsequent gravitational-assisted descent during inspiration.27 There may beinward movements of the lower rib cage during inspiration (Hoover's sign).These clinical features are seldom present until diaphragm strength is reduced toa quarter of normal, thus substantial diaphragm weakness can be overlooked onclinical examination.'20

Radiology and imaging

The radiological assessment of diaphragmatic function has traditionally reliedupon plain radiography and erect fluoroscopy.28 In hemidiaphragm paralysis, anelevated hemidiaphragm may be visible on plain chest X-ray. Similarly, in bilat-eral diaphragm paralysis, chest X-ray may show elevation of both domes of thediaphragm. Unilateral or bilateral diaphragmatic paralysis may also be detectedby radiological fluoroscopic screening, with the patient in a supine position andtaking short sharp sniffs.' 27 Normally, the left diaphragm moves more than theright, with a craniocaudal excursion of 3 and 6 cm, respectively. Paradoxicalmovement occurs when the pressure exerted by the abdomen exceeds that of thediaphragm so that movement is upward on inspiration or sniffing. This can occurin diaphragm paralysis due to injury or tumour involvement of the phrenic nerves.Determining whether diaphragm elevation is due to paralysis or to an

abdominal mass may be difficult. Ultrasound is the definitive diagnostic investi-gation. Axial movements of the right hemidiaphragm during tidal breathing canalso be recorded using real-time ultrasonography.29 30 This investigation may beof clinical value in the assessment ofdiaphragm paralysis,30 diaphragm rupture,3rib cage abnormalities, and abdominal defects in newborns.29

Volitional tests

PULMONARY FUNCTION TESTSEvaluation of pulmonary function would be incomplete without assessment ofthe respiratory muscles."6 Weakness of both inspiratory and expiratory musclesreduces vital capacity (VC). However, a fall in VC occurs only when weakness ismoderate to severe, therefore, a normal VC excludes clinically significant respi-ratory muscle weakness.26 The accuracy of VC measurement is dependent oneffort and co-operation by the patient. A major limitation of VC is that it isreduced by factors other than weakness, therefore, a low VC can be non-specificand non-diagnostic.26 Furthermore, VC is reduced by many pulmonary, cardio-vascular and systemic diseases. Thus, it can be a rather insensitive, inaccurateand non-specific measurement of respiratory muscle weakness.'

In patients with substantial diaphragm weakness, VC is decreased in thesupine position; the fall in supine VC is more prominent in patients who haveorthopnoea and abdominal paradox." In patients with bilateral diaphragmparalysis, the decline in VC is approximately 50%.'9 In a clinical setting, VC is ofgreat value in monitoring patients at risk of rapidly progressive muscle weaknessor paralysis, for example, Guillain-Barre syndrome."Other pulmonary function tests for abnormalities due to respiratory muscle

weakness include: reduced total lung capacity and functional residual capacity,usually with a normal residual volume. The restrictive ventilatory defect causedby respiratory muscle weakness reduces overall gas exchange. The transfer fac-tor for carbon monoxide (TLco) is reduced but it is normal when adjusted forvolume. This is important in distinguishing respiratory muscle weakness fromlung parenchymal disorders, such as pulmonary fibrosis and pulmonary oedema.Pulmonary function test abnormalities in patients with severe respiratory mus-cle weakness are shown in box 4.

ARTERIAL BLOOD GASESIsolated bilateral diaphragm paralysis rarely causes ventilatory failure.'9 Lowarterial partial pressure of oxygen (PaO2) and/or high arterial partial pressure ofcarbon dioxide (PaCO2) will not normally develop until respiratory musclestrength is reduced to less than 30% of normal.' Initially, hypoxaemia andhypercapnia occur only during sleep, and sleep studies are therefore a part ofassessment. From a clinical standpoint, it is important to note that the presenceof hypercapnia implies severe weakness, and hypercapnia when awake is usuallyassociated with a poor prognosis.

POLYSOMNOGRAPHYSleep studies including oximetry may be useful in detecting nocturnal desatura-tion due to respiratory muscle weakness. A normal nocturnal saturationobviously excludes severe respiratory muscle weakness or paralysis. Althoughnocturnal desaturation does not occur until global respiratory muscle strength is

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from

210 Syabbalo

Metabolic and endocrinedisorders which predispose torespiratory muscle weakness

Malnutrition* beriberi* alcoholic myopathy* anorexia nervosa

Metabolic* hypercapnia* hypoxaemia* chronic renal insufficiency* hypocalcaemia* hypokalaemia* hypomagnesaemia* hypophosphataemia* periodic paralysis* acute intermittent porphyria

Endocrine* diabetes mellitus* steroid myopathy* hypothyroidism* thyrotoxicosis* hyperparathyroidism* adrenal insufficiency* acid maltase deficiency

Box 2

Indications for measurementsof respiratory muscle strength

* investigation of patients withunexplained dyspnoea and cough

* evaluation of respiratory musclestrength in patients withneuromuscular diseases

* evaluation of respiratory muscle powerin patients with cardiopulmonarydiseases

* ventilator weaning protocols* evaluation of therapeutic interventions

(follow-up of patients withneuromuscular and cardiopulmonarydiseases, patients on corticosteroidtherapy, and patients on inspiratorymuscle training)

Box 3

reduced to below one third of normal in patients with neuromuscular diseasealone, it may occur earlier if there is coexisting respiratory disease, such aschronic obstructive pulmonary disease and interstitial lung disease.

In some patients, severe isolated diaphragm paralysis is associated withnocturnal desaturation, particularly during rapid eye movement sleep, and theimportance of this has yet to be elucidated.'9 24 Collapse of the upper airwayduring sleep may aggravate the situation and is relatively more common inpatients with generalised neuromuscular disease.26

LIMB MUSCLESWhen assessing patients with respiratory muscle weakness it is useful to assessthe strength of the peripheral skeletal muscles, since weakness of these musclesis frequently evident in neuromuscular and systemic diseases.2" For example, thehandgrip force is reduced in patients with neuromuscular disease, and in patientswith cardiac cachexia due to congestive heart failure.'0

MOUTH PRESSURESThe most widely used test of global inspiratory and expiratory muscle strengthare the static maximum pressures measured at the mouth (PImax andPEmax).'6 "333 These tests have the advantage that they are noninvasive and nor-mal values have been established in both adults""-5 and children."6 A high PImax(80 cmH2O) or a PEmax (90 cmH2O) excludes clinically important inspiratory26or expiratory weakness. PImax and PEmax are influenced by age," gender,l""Sposture, lung volume and the type of mouthpiece.'7 They are preferablymeasured in the sitting position using a standard flanged mouth piece.'6Conventionally, PImax is measured from residual volume, and PEmax from totallung capacity, with a noseclip. The highest recorded pressure maintained for 1 srepresents PImax or PEmax. Portable mouth pressure metres allow immediatemeasurement of PImax and PEmax at the bedside or in the clinic.'6 PImax canbe easily measured during routine follow-up of ambulatory patients with chronicobstructive pulmonary disease. It is also useful in assessing respiratory musclestrength in patients with respiratory failure who cannot undergo an invasiveassessment of inspiratory muscle due to respiratory distress.'8There are a few drawbacks to measurements of PImax and PEmax. One

problem with static pressures is that they are volitional, thus, for the result to bea true measure of strength, the subject must make maximal contraction.26 39 Fur-thermore, there is wide inter-subject variability in PImax and PEmax in normalindividuals, and significant variation in reported normal values for PImax.40 Thismay be due to subject motivation, number of pre-measurement runs (learningeffect), and peri-oral leaks, especially at higher pressures and in older subjects.The PImax manoeuvre is both demanding and unpleasant.4' Low PImax mayalso be due to lack of motivation or co-ordination.

Clinically, PImax is the most widely applied technique in assessment ofinspiratory muscle strength." PEmax is the only test generally available formeasurement of expiratory muscle power.'6

SNIFF NASAL PRESSUREThere is a close correlation between nasal, nasopharyngeal, mouth, andoesophageal pressure during a sniff in normal subjects,4""3 and in patients with-out severe nasal disease. A useful alternative method for measuring inspiratorymuscle strength consists of asking the subject to perform short sniffs ofmaximalintensity.4' The sniffmanoeuvre is natural and probably easier to perform thanthe PImax manoeuvre for most subjects.4' 44 Thus, inspiratory muscle strength isoften better reflected by maximal sniff pressure (SNIP) than by the PImaxmanoeuvre.4' SNIP is measured through a plug occluding one nostril duringsniffs performed through the other nostril.4' Nasal leaks are eliminated by usingwaxed ear plugs hand-fastened around the tip of a polyethylene catheter.4' SNIPis measured from functional residual capacity rather than from residual volume,because inspiratory muscle strength is overestimated at levels below functionalresidual capacity due to the elastic recoil pressure of the thorax.4' In clinicalpractice, SNIP is easier to measure than sniff nasopharyngeal pressure (sniffPnp). SNIP can be predicted from age by a first degree equation for both sexes.Normal values of SNIP have been established in one large study of 160 subjects(men> 70 cmH2O, women> 60 cmH2O) A

Nonetheless, the sniffmanoeuvre is volition-dependent, and submaximal effortsare most likely to be recorded in patients who are ill or breathless.46 Other limita-tions of SNIP include anatomical abnormalities such as nasal polyps, nasal septaldefects, and septal deviation, that may impair pressure transmission from the rhi-nopharynx. Nasal mucosal congestion in patients with rhinitis predisposes to thesame kind of limitation.4' In the described conditions, there is underestimation

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from

Assessment of respiratory muscle function 211

Pulmonary function tests inpatients with severerespiratory muscle weakness

Vital capacity (% predicted): reducedFEVI/FVC (% predicted): normalPEF (% predicted): reducedFEF25 50 (% predicted): reducedResidual volume (% predicted): normal

or decreasedFunctional residual volume (%

predicted): decreasedTLcc (% predicted): reducedTLCO/VA (% predicted): normalAirway resistance: normalAbbreviations as in the text.

Box 4

of sniff oesophageal pressure by SNIP. Furthermore, SNIP measurements areunreliable in the presence of airway obstruction or pulmonary fibrosis that preventaccurate transmission of pleural pressure to the upper airways.26 In patients withchronic obstructive pulmonary disease the sniff inspiratory pressure underesti-mates inspiratory muscle strength,47 because the sniff is a short manoeuvre.48However, SNIP is very useful in the evaluation and follow-up ofpatients with neu-romuscular or skeletal disorders.Maximal SNIP is similar in the sitting and supine position and is significantly

higher than PImax in healthy subjects45 for several reasons:* the sniff measurement is easier and less unpleasant than the PImaxmanoeuvre and may thereby allow maximal muscular recruitment to beachieved more often45

* dynamic changes in human diaphragm length during maximal inspiratoryeffort against occlusion using sequential radiography, indicate that thismanoeuvre does not represent an isometric contraction

* the level of recruitment of inspiratory muscle group is different during sniffand during PImax manoeuvre.SNIP and PImax are not interchangeable but rather compliment one another

in the assessment of inspiratory muscle strength. In patients suspected of respi-ratory muscle weakness, a SNIP may be required to confirm respiratory muscleweakness ifPImax is low in order to rule out lack of motivation or co-ordinationduring the PImax manoeuvre. The development of the SNIP test represents afurther step towards accurate, easy, and noninvasive assessment of inspiratorymuscle strength.26

OESOPHAGEAL, GASTRIC, AND TRANSDIAPHRAGMATIC PRESSUREIn some patients, simple noninvasive tests will fail to confirm or refute thepossibility of respiratory muscle weakness. Similarly, for patients with knownweakness it is important to make a precise estimation of the weakness in order tomake decisions about management. Placement of oesophageal and gastricballoons using topical anaesthesia allow measurement of oesophageal, gastric,and transdiaphragmatic pressure, and also dynamic compliance. Oesophagealand gastric pressure are measured by two balloons, one placed in themid-oesophagus (Poes) and the other in the stomach (Pga) during a PImaxmanoeuvre. The balloons should be positioned and tested in the standardmanner.49 Oesophageal pressure reflects pleural pressure, and Pga reflectsabdominal pressure. Transdiaphragmatic pressure (Pdi) is the pressuredifference between Pga and Poes. Pdi reflects the tension developed by the dia-phragm, and is viewed as a reflection of diaphragmatic inspiratory action.9 Thesemeasurements depend on lung volumes and they should be standardised for thelung volumes at which they are measured.The measurement of Pdi using two balloon catheters is not an easy or short

procedure, and in some patients it may not be well-tolerated. Although the two-balloon technique is an invasive technique, PImax measured during voluntarymanoeuvres is not a reliable index of diaphragm function. Furthermore, PImaxdoes not permit discrimination between various inspiratory muscle groups.Therefore, measurement of Pdi may be required to assess diaphragmatic func-tion specifically.

SNIFF OESOPHAGEAL AND TRANSDIAPHRAGMATIC PRESSURESIn the search for a reliable and practicable test of inspiratory muscle strength,considerable interest has recently focused on the sniff manoeuvre.42 50 51 Inspira-tory muscle strength is often better reflected by oesophageal pressure during amaximal sniff (sniff Poes).42 Similar to SNIP, sniff Poes is performed from func-tional residual capacity without a noseclip. When thus performed the volumeincrease is about 500 ml,52 and diaphragm contraction is therefore relatively iso-metric. The normal mean sniff Poes is 93 ± 20 cmH2O, range 74-135 cmH2O.The mean value for the ratio SNIP/sniff Poes is 0.91 (0.82-0.99). In normalsubjects sniff Poes and sniff Pdi have a narrower normal range and a betterbetween-day coefficient variation than the PImax manoeuvre.44 Occasionally,there is overestimation of sniff Poes because of agonist activation of the orofacialmuscles in some normal subjects.5'

Sniff Poes also provides a reliable estimation of inspiratory muscle strength inpatients with neuromuscular and skeletal disorders.54 However, in patients withaltered lung mechanics, sniff Poes is usually lower than SNIP. Recently, Uldryand colleagues54 have shown that the mean SNIP/sniff Poes ratio in patients withchronic obstructive pulmonary disease is 0.80, compared with 0.90 in normalsubjects.4' SNIP underestimates inspiratory muscle strength due to impairedtransmission of pleural pressure in the presence of airway obstruction.46 54 Thisdifference could be expected in view of the short and dynamic character of the

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from

Syabbalo

sniff.54 Other limitations of sniff Poes and sniff Pdi measurements include ana-tomical nasal abnormalities, eg, nasal polyps, septal deviation, septal defects, andnasal mucosal congestion due to rhinitis, which may impair pressuretransmission from the rhinopharynx.4'The validity of sniff testing also depends on the subject making a maximal

effort. It is increasingly recognised that submaximal activation is common dur-ing volitional muscle testing, particularly in patients who are ill or breathless.46Nevertheless, in clinical practice, sniff Poes and sniff Pdi are the most accurateand reproducible volitional tests currently available to assess global inspiratory,and diaphragmatic strength.26Non-volitional tests

PHRENIC NERVE ELECTRICAL STIMULATIONThe twitch occlusion technique was endorsed by the 1989 National Heart, Lungand Blood Institute workshop on respiratory muscle fatigue as a potentiallymajor diagnostic tool.55 Currently, twitch oesophageal pressure (TwPoes) hasbeen used to assess respiratory muscle function in normal subjects,56 and inpatients with respiratory muscle weakness.34 57 58 Bilateral electrical phrenicnerve stimulation (EPS) in the neck and the measurement oftwitch Pdi (TwPdi)has the advantage of eliminating variation due to patient motivation. It alsoallows measurement ofelectromyographic signals and nerve conduction (normal(SD) range 7.7 (0.8) ms).59 60 In addition, hemidiaphragm function may be suc-cessfully assessed by unilateral electrical stimulation of the phrenic nerves.'8Twitch potentiation is a potential problem during assessment of diaphragmstrength, irrespective ofthe method of stimulation.6 62 After a maximal voluntarycontraction of the diaphragm, the TwPdi at functional residual capacity canincrease by 70%.61 To minimise potentiation, subjects should rest, breathingquietly for 10-20 minutes before testing.6' The normal range for TwPdi is 8.8-33.0 cmH,O. This broad range makes it diagnostically useful only when there issevere weakness or when the result is equivocally normal.58A major drawback to EPS is that it is not an easy technique to perform, and

the operator must be highly trained for the results to be relied upon.9 Currentintensities required to achieve supramaximal activation can exceed 4-50 mA,and can be perceived as too painful by some subjects. Patients may find itdifficult to relax during EPS, thereby causing an increase in twitch tension - thephenomenon of twitch potentiation.6' 62 Furthermore, EPS is technically diffi-cult in obese patients and Cushingoid patients with "buffalo-hump".63 Indeedthe phrenic nerve may be impossible to locate. EPS is potentially dangerous ininfants and small children,64 and should be avoided in patients with epilepticseizures. Despite some promising findings, this method has yet to gainwidespread clinical acceptance because oftechnical concerns relating to patientdiscomfort, stimulation of the muscle groups, and concerns about the vari-ability of phrenic nerve stimulation.

EPS Twitch mouth pressureIt is also possible to measure twitch Pmo (TwPmo) during EPS.56 TwPmo cor-relates well with TwPoes, at relaxed functional residual capacity and relaxed vol-umes above functional residual capacity in normal subjects. However, in patientswith chronic obstructive airways disease, pressure equilibration between alveoliand mouth is incomplete due to the airways acting as a resistance-capacitancefilter between the two. Because the airway time is long and diaphragmatic con-traction time short, TwPmo appears damped and time-lagged with respect toTwPoes.9

MAGNETIC PHRENIC NERVE STIMULATION

By discharging a magnetic coil, it is possible to create a pulsed magnetic field thatcauses current to flow in nervous tissue, which in turn causes muscle tocontract.65 Cervical magnetic stimulation (CMS) is performed using a magneticstimulator (Magstim 200, The Magstim Company Limited, Whitland,Carmarthenshire, Wales). Supramaximal bilateral phrenic nerve stimulationmay be achieved using a 9-10 cm circular coil placed over the cervical phrenicnerve roots at the spinous processes of the 5th to 7th cervical vertebra.66 67 CMSshould be performed at functional residual capacity, after airway occlusion, withthe diaphragm relaxed. Mean (SD) TwPdi for normal subjects is 31 (6)cmH20.63 The magnetically elicited TwPdi is slightly greater than the electrical,the difference being due to a larger TwPoes.67 68 Magnetic stimulation has ahigher and better defined lower limit of normal than EPS, and a better correla-tion with sniff Pdi, which increases its diagnostic sensitivity in patients withmoderate diaphragm weakness,69 and allows sequential studies in patients.70

212 on 17 July 2018 by guest. P

rotected by copyright.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.74.870.208 on 1 April 1998. D

ownloaded from

Assessment of respiratory muscle function

During CMS, great care should be taken to locate the optimal position for themagnetic coil on the neck, and the neck should be flexed to obtain maximalcontact.63 Failure to achieve this may result in significantly submaximal stimula-tion. This can occur if the coil cannot be placed close enough to the phrenicnerves, eg, in patients with ankylosing spondylitis, systemic lupus erythemato-sus, or Cushing syndrome with 'buffalo-hump'.63 Stimulation of the phrenicnerves anteriorly with a figure of eight coil allows easy and accurate quantifica-tion of diaphragm function in supine subjects,7 72 for example, patients in theintensive care unit,26 71 or undergoing coronary artery surgery.73CMS is easier to perform than EPS, is painless and perhaps 'closer to life,'

because it stabilises the rib cage. It is possible to achieve stimulation without theproblem of twitch potentiation. Reproducibility is slightly better than for EPS.67In clinical practice, the TwPdi invoked by this method is a valid measure of dia-phragm strength that is acceptable to patients.63 The major advantage ofmagnetic stimulation is that the test is independent of patient aptitude andmotivation. Currently, CMS is beginning to be used in a clinical setting, includ-ing evaluating difficult weaning from mechanical ventilation after cardiacsurgery.74 It is likely to be of value in investigation of low frequency fatigue inclinical studies.70 Because patients find it so acceptable, it may be a useful tech-nique for evaluating strategies aimed at altering respiratory muscle strength, ie,inspiratory muscle training.

CMS Twitch mouth pressureMeasuring pressure at the mouth during magnetic stimulation of the phrenicnerves can be used to assess diaphragm contractility.75 In normal subjects, meanTwPmo is 13.7 cmHO (range 11.3-16.1) and TwPoes is 13.3 cmH20 (range10.4-15.9). In normal subjects, the relationship between TwPmo and TwPoes isclose to unity with a mean (SD) difference of 0.1 (1.13) cmH20.75 Preliminarystudies have shown that TwPmo measured by CMS reflects TwPoes and TwPdiin some patients with lung disease.75 An advantage ofthis method is that it avoidsthe need for oesophageal and gastric balloons, which are usually consideredunpleasant by stable patients.76ELECTRICAL VS MAGNETIC STIMULATIONElectrical stimulation is often uncomfortable, particularly when attempts aremade to achieve bilateral supramaximal stimulation,67 while localising thephrenic nerves can be difficult in some subjects with short necks, 'buffalo-hump', and kyphoscoliosis. Because maintaining a constant symmetrical maxi-mal stimulus can be difficult, repeated twitch stimulations are often performed,which in itself can increase twitch pressure, the so-called staircasephenomenon.6 62 It has been shown that maximal voluntary contraction canincrease TwPdi by up to 70%.61 With CMS, locating the phrenic nerve roots isnot a much of a technical problem, and intact nerves are stimulated invariably.This technique is considerably less painful than EPS, and the risk ofthe staircasephenomenon is low because locating the area of maximal stimulation usuallyrequires only a few stimulations. In normal subjects at relaxed functional residualcapacity, TwPdi elicited by CMS is greater than that elicited by EPS.67 77 78 Thereason for the discrepancy in TwPdi values with the two techniques is not com-pletely understood. Magnetic stimulation is considered to act on both the cervi-cal roots and phrenic nerve trunks,66 and thus should not be affected by thepresence of ectopic branches of the phrenic nerve. In addition, contractions ofthe sternocleidomastoid, trapezius, parasternal and pectoralis muscles areknown to decrease rib cage compliance, which should result in a larger TwPoesfor a given diaphragmatic contraction.78 Recruitment of the extradiaphragmaticmusculature by magnetic stimulation is possibly responsible for the difference inregional motion of the rib cage between the two stimulation techniques.

In EPS the stimulation is relatively confined to the diaphragm, while CMSstimulates the diaphragm and muscles of the upper rib cage, which may cause asomewhat higher magnetic than electrical TwPdi.67 However, the relativenon-specificity of cervical magnetic stimulation does not undermine its value inthe assessment of diaphragmatic function. CMS is as effective as EPS in thedetection of diaphragm weakness or fatigue.MAGNET TICSTIMULATION OF THE CORTEX

Recently, it has been shown that respiratory muscle weakness can be assessedusing cortical stimulation.7779 Magnetic stimulation of the cortex (CxMS) isperformed with a Magstim 200. The coil is positioned over the vertex, 1 to 3 cmbehind the midauricular plane,77 at maximal stimulation intensity. In order tominimise twitch potentiation, the subject should breathe quietly and remainrelaxed for 20 min.61 CMS and CxMS are performed at functional residual

213 on 17 July 2018 by guest. P

rotected by copyright.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.74.870.208 on 1 April 1998. D

ownloaded from

214 Syabbalo

Step-wise assessment ofpatients with respiratorymuscle weakness

* history and clinical examination* chest imaging* pulmonary function tests: sitting and

supine vital capacity, transfer factor,arterial blood gases

* maximal inspiratory and expiratorypressures (PImax and PEmax)

* measurement of Poes, Pga, and Pdiduring PImax manoeuvre

* sniff inspiratory pressures (sniff Pmo,sniff Pnp, sniff Poes, sniff Pdi)

* cervical magnetic stimulation ofphrenic nerves (TwPmo, TwPoes,TwPdi)

* electrical stimulation of the phrenicnerves (TwPmo, TwPoes, TwPdi)

* cervical magnetic stimulation ofphrenic nerves (phonomyogram)

* cortical magnetic stimulation(TwPoes, TwPdi)

Abbreviations as in text

Box S

capacity, after airway occlusion, with the diaphragm relaxed. TwPoes and TwPdiare produced in a similar manner as during CMS. CxMS is contraindicated inpatients with epileptic seizures, intracranial lesions, and heart pacemakers.79 Infuture, CxMS could prove to be useful in the assessment of the central drive torespiratory muscles, and in elucidating the mechanisms of central respiratorymuscle fatigue. In addition, it may provide information that may help to explainthe differences in Pdi observed with CMS and EPS,77 and the difference betweencentral and peripheral fatigue.

PHONOMYOGRAPHYWhen skeletal muscles contract, they produce vibrations and low frequencysounds easily recorded via a microphone taped on the skin.80 This mechanicalsignal is called a phonomyogram. The phonomyogram provides an index of iso-metric individual muscle force and motor recruitment.8' During electricalstimulation, there is a significant correlation between the phonomyogram signaland current pulse intensity. As for the surface electromyogram signal, the surfacephonomyogram signal recorded during spontaneous breathing represents asummation of various muscle activities. Diaphragm function may similarly beassessed directly by recording the phonomyogram during EPS and CMS. Thissound signal, which can be recorded noninvasively with surface microphones, isdirectly proportional to the tension developed during contraction. Measurementof the sound signal produced by contraction of the diaphragm (phonomyogra-phy) appears attractive,82 because this method is less invasive and does notrequire insertion of oesophageal balloons. Although this technique is still understudy, it may well become the standard method for investigating respiratorymuscle strength.76

Conclusions

A scheme for the step-wise assessment of respiratory muscles weakness is shownin box 5. However, the most attractive and innovate assessment of diaphragmfunction include recording the phonomyogram during magnetic stimulation.The combination of an easy and painless magnetic stimulation and phonomyog-raphy may be the standard method for assessment ofrespiratory muscle functionin the next millennium.

1 Green M, Moxham J. Respiratory muscle inhealth and disease. In: Barnes P, ed, Respiratorymedicine: recent advances. Oxford: Butterworth-Heinemann, 1993; pp 252-75.

2 Braun N, Arora N, Rochester D. Respiratorymuscle and pulmonary function in poliomyelitisand other proximal myopathies. Thorax 1983;38:616-23.

3 Laroche CM, Mier AK, Spiro SG, Newsom-Davis J, Moxham J, Green M. Respiratory mus-cle weakness in the Lambert-Eaton myasthenicsyndrome. Thorax 1989;44:13-8.

4 Mier A, Brophy C, Moxham J, Green M.Repetitive stimulation of phrenic nerves inmyasthenia gravis. Thorax 1992;47:640-4.

5 Keenan SP, Alexander D, Road JD, Ryan CF,Oger J, Wilcox PG. Ventilatory muscle strengthand endurance in myasthenia gravis. Eur RespirJ 1995;8:1130-5.

6 Murciano D, Rigaud D, Pingleton S, Armen-gaud M, Melchior J, Aubier M. Diaphragmaticfunction in severely malnourished patients withanorexia nervosa. Am J Respir Crit Care Med1994;150: 1569-74.

7 Rochester DF, Braun NMT Determinants ofmaximal inspiratory pressure in chronic ob-structive pulmonary disease. Am Rev Respir Dis1985;132:42-7.

8 Corrado A, Bruscoli G, De Paola E, Ciardi-Dupre GF, Baccini A, Taddei M. Respiratorymuscle insufficiency in acute respiratory failureof subjects with COPD: treatment with inter-mittent negative pressure ventilation. Eur RespirJ 1990;3:644-8.

9 Similowiski T, Derenne JPH. Inspiratory muscletesting in stable COPD patients. Eur Respir J1994;7:1871-6.

10 Hammond MD, Bauer KA, Sharp JT, RochaRD. Respiratory muscle strength in congestiveheart failure. Chest 1990;98:1091-4.

11 McParland C, Krishnan B, Wang Y, GallagherCG. Inspiratory muscle weakness and dyspneain chronic heart failure. Am Rev Respir Dis1992;148:467-72.

12 Ambrosino N, Opasich C, Crotti P, Cobelli F,Tavazzi L, Rampulla C. Breathing pattern, ven-tilatory drive and respiratory muscle strength inpatients with chronic heart failure. Eur Respir J1994;7: 17-22.

13 Coakley JH, Nagendran K, Honavar M, HindCJ. Preliminary observations on the neuromus-cular abnormalities in patients with organfailure and sepsis. Intensive Care Med 1993;19:323-8.

14 Juan G, Calverley P, Talam C, Schnader J, Ros-sous C. Effect of CO2 on diaphragm function innormal human beings. NEnglJMed 1984;310:874-9.

15 Decramer M, Lacquet L, Fagaud R, Rogiers P.Corticosteroids contribute to muscle weaknessin chronic airflow obstruction. Am Jf Respir CritCare Med 1994;150:1 1-6.

16 Hamnegard C-H, Wragg S, Kyroussis D, Aqui-lina R, Moxham J, Green M. Portablemeasurement of maximum mouth pressures.Eur RespirJ 1994;7:398-401.

17 Moxham J, Goldstone J. Assessment of respira-tory muscle strength in the intensive care unit.Eur RespirJ 1994;7:2057-6 1.

18 Laroche CM, Mier AK, Moxham J, Green M.Diaphragm strength in patients with recenthemidiaphragm paralysis. Thorax 1988;43: 170-4.

19 Laroche CM, Carroll N, Moxham J, Green M.Clinical significance of severe isolated dia-phragm weakness. Am Rev Respir Dis 1988;138:862-6.

20 Moxham J. Respiratory muscle fatigue: mecha-nisms, evaluation and therapy. Br J7 Anaesth1990;65:43-53.

21 Mier-Jedrzejowicz A, Brophy C, Moxham J,Green M. Assessment of diaphragm weakness.Am Rev RespirDis 1988;137:877-83.

22 Davison A, Mulvey D. Idiopathic diaphragmaticweakness. BMJ 1992;304:492-4.

23 Fromm GB, Wisdom PJ, Block AJ. Amyo-trophic lateral sclerosis presenting with respira-tory failure. Chest 1977;71:612-4.

24 White J, Drinnan M, Smithson A, Griffiths C,Gibson G. Respiratory muscle activity and oxy-genation during sleep in patients with muscleweakness. Eur RespirJ 1995;8:807-14.

25 Smith PE, Calverley PM, Edwards RH. Hypox-emia during sleep in Duchenne musculardystrophy. Am Rev Respir Dis 1988;137:884-8.

26 Polkey MI, Green M, Moxham J. Measurementof respiratory muscle strength. Thorax 1995;50:1131-5.

27 Alexander C. Diaphragmatic movements andthe diagnosis of diaphragmatic paralysis. ClinRadiol 1966;17:79-83.

28 Flo wer CDR. The diaphragm. Diagnostic radiol-ogy, vol 1. London: Churchill Livingstone,1986; p170.

29 Laing IA, Teele RL, Stark AR. Diaphragmaticmovement in new-born infants. J Pediatr 1988;112:638-43.

30 Houston JG, Angus RM, Cowan MD, McMil-lan NC, Thomson NC. Ultrasound assessmentof normal hemidiaphragmatic movement: rela-tion to inspiratory volume. Thorax 1994;49:500-3.

31 Somers JM, Gleeson FV, Flower CDR. Ruptureof the diaphragm following blunt trauma: theuse of ultrasound in diagnosis. Clin Radiol 1990;42:97-101.

32 Chevrolet J, Delamont P. Repeated vital capac-ity measurements as predictive parameters formechanical ventilation need and weaning suc-cess in Guillain-Barre syndrome. Am Rev RespirDis 1991;144:814-8.

33 Black LF, Hyatt RE. Maximal respiratorypressures: normal values and relationships toage and sex. Am Rev Respir Dis 1969;99:696-702.

34 Leech J, Ghezzo H, Stevens D, Becklake M.Respiratory pressures and function in youngadults. Am Rev Respir Dis 1983;128:17-23.

35 Bruschi C, Cerveri I, Zoia MC, et al. Referencevalues of maximal inspiratory mouth pressures:a population based study. Am Rev Respir Dis1992;146:790-93.

36 Gauthier C, Zinman R. Maximal static pres-sures in children. Respir Physiol 1983;51:45-6 1.

37 Koulouris N, Mulvey DA, Laroche CM, GreenM, Moxham J. Comparison of two differentmouthpieces for the assessment of PImax andPEmax in normal and weak subjects. Eur RespirJ 1988;1:863-7.

38 Wijkstra PJ, van der Mark TW, Boezen M, vanAltena R, Postma DS, Koeter G. Peak inspira-tory mouth pressure in healthy subjects and inpatients with COPD. Chest 1995;107:652-5.

39 Aldrich T, Spiro P. Maximal inspiratorypressure: does reproducibility indicate fulleffort. Thorax 1995;50:40-3.

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from

Assessment of respiratory muscle function 215

40 Smyth RJ, Chapman KR, Rebuck SR. Maximalinspiratory and expiratory pressures inadolescents: normal values. Chest 1986;4:568-72.

41 Heritier F, Rahm F, Pasche P, Fitting J-W. Sniffnasal pressure. A noninvasive assessment ofinspiratory muscle strength. Am J Respir CritCare Med 1994;150:1678-83.

42 Koulouris N, Mulvey DA, Laroche CM, Saw-icka EH, Green M, Moxham J. Themeasurement of inspiratory muscle strength bysniff esophageal, nasopharyngeal, and mouthpressure. Am Rev Respir Dis 1989;139:641-6.

43 Kyroussis D, Mills G, Hamnegard C-H, et al.Inspiratory muscle relaxation rate assessed fromsniff nasal pressure. Thorax 1994;49:1127-33.

44 Miller JM, Moxham J, Green M. The maximalsniff in the measurement of diaphragm functionin man. Clin Sci 1985;69:91-6.

45 Uldry C, Fitting J-W. Maximal values of sniffnasal inspiratory pressure in healthy subjects.Thorax 1995;50:371-5.

46 Mulvey DA, Elliot MW, Coulouris NG, et al.Sniff, esophageal and nasopharyngeal pressuresand maximal relaxation rates in patients withrespiratory dysfunction. Am Rev Respir Dis1991;143:950-3.

47 Uldry C, Fitting J. Influence of airway obstruc-tion on sniff nasal inspiratory pressure. Am JfRespir Crit Care Med 1995;151:A414.

48 Murciano D, Aubier M, Bussi S, Derenne JP,Pariente R, Milic-Emili J. Comparison ofesophageal, tracheal, and mouth occlusion pres-sure in patients with chronic obstructive pulmo-nary disease during acute respiratory failure.Am Rev Respir Dis 1982;126:837-41.

49 Milic-Emili J, Mead J, Turner JM, Glauser EM.Improved technique for estimating pleural pres-sure from esophageal balloons. Jf Appl Physiol1964;19:207-1 1.

50 Laroche CM, Mier AK, Moxham J, Green M.The value of sniff esophageal pressures in theassessment of global inspiratory musclestrength. Am Rev Respir Dis 1988;138:598-603.

51 Mulvey DA, Elliot MW, Coulouris NG, et al.Sniff, esophageal and nasopharyngeal pressuresand maximal relaxation rates in patients withrespiratory dysfunction. Am Rev Respir Dis1991;143:950-3.

52 Pertuze J, Watson A, Pride N. Limitation ofmaximum inspiratory flow through the nose.Bull Eur Physiol Respir 1987;23 (suppl 12):344s.

53 Heritier F, Rahm F, Pasche P, Fitting JW. Sniffnasal inspiratory pressure: a noninvasive assessment of inspiratory muscle strength. Am _J RespirCrit Care Med 1994;150:1678-83.

54 Uldry C, Janssens JP, de Muralt B, Fitting JW.Sniff nasal inspiratory pressure in patients withchronic obstructive pulmonary disease. EurRespirj 1997;10:1292-6.

55 NHLBI Workshop on respiratory muscle fa-tigue report. Am Rev Respir Dis 1990;142:474-80.

56 Yan S, Gauthier AP, Similowski T, MacklemPT, Bellemare F. Evaluation of human dia-phragm contractility using mouth pressuretwitches. Am Rev Respir Dis 1992;145:1064-9.

57 Mier A, Brophy C, Moxham J, Green M.Phrenic nerve stimulation in normal subjectsand in patients with diaphragmatic weakness.Thorax 1987;42:885-8.

58 Mier A, Brophy C, Moxham J, Green M. Twitchpressures in the assessment of diaphragmaticweakness. Thorax 1989;44:990-6.

59 Newsom-Davis J. Phrenic nerve conduction inman. J Neurol Neurosurg Psychiatry 1967;30:420-5.

60 Chen R, Collins S, Remtulla H, Parkes A, Bot-ton C. Phrenic nerve conduction study innormal subjects. Muscle Nerve 1995;18:330-5.

61 Wragg S, Hamnegard C, Road J, et al. Potentia-tion of diaphragmatic twitch after voluntarycontraction in normal subjects. Thorax 1994;49:1234-7.

62 Mador MJ, Magalang UJ, Kufel TJ. Twitchpotentiation following voluntary diaphragmaticcontraction. Am J Respir Crit Care Med 1994;149:739-43.

63 Hamnegard C-H, Wragg SD, Mills GH, et al.Clinical assessment of diaphragm strength bycervical magnetic stimulation of the phrenicnerves. Thorax 1996;51:1239-42.

64 Gauthier C. Respiratory muscle function ininfants. EurRespirJ 1995;8:150-3.

65 Machetanz J, Bischoff C, Pilchlmeier R, MeyerB-U, Sader A, Conrad B. Magnetically induced-muscle contraction caused by motor nervestimulation and not by direct muscle activation.Muscle Nerve 1994;17:1170-5.

66 Similowiski T, Fleury B, Launois S, Cathala HP,Bouche P, Derenne JP. Cervical magneticstimulation: a new painless method for bilateralphrenic nerve stimulation in conscious humans.J Appl Physiol 1989;67:1311-8.

67 Wragg S, Aquilina R, Moran J, et al. A compari-son of cervical magnetic stimulation and bilat-eral percutaneous electrical stimulation of thephrenic nerves in normal subjects. Eur Respir J1994;7:1788-92.

68 Similowiski T, Duguet A, Straus C, Attali V.Boisteanu D, Derenne J-Ph. Assessment of vol-untary activation of the diaphragm usingcervical and cortical magnetic stimulation. EurRespirJ 1996;9:1224-1231.

69 Wragg S, Hamnegard C-H, Kyroussis D, MillsG, Green M, Moxham J. Assessment ofdiaphragm strength in patients using cervicalmagnetic stimulation. AmJRespir Crit Care Med1994;149:A130.

70 Polkey MI, Kyroussis D, Keilty SEJ, et al.Exhaustive treadmill exercise does not reducetwitch transdiaphragmatic pressure in patientswith COPD. Am Jf Respir Crit Care Med1995;151:959-64.

71 Mills GH, Kyroussis D, Hamnegard CH, WraggS, Moxham J, Green M. Unilateral magneticphrenic nerve stimulation in unilateral dia-phragm paresis. Thorax 1994;49: 1058-9P.

72 Mills GH, Kyroussis D, Hamnegard CH, PolkeyM, Green M, Moxham J. Bilateral magneticstimulation of the phrenic nerves using an ante-rior approach. Thorax 1994;49: 1058P.

73 Mills GH, Ponte J, Hamnegard CH, et al.Human diaphragm contractility during intraop-erative cooling. AmJRespir Crit Care Med 1995;151:A809.

74 Diehl JL, Lofaso F, Deleuze P, Similowiski T,Lemaire F, Brochard L. Clinically relevantdiaphragmatic dysfunction after cardiac opera-tions. Jf Thorac Cardiovasc Surg 1994;107:487-98.

75 Hamnegard C-H, Wragg S, Kyroussis D, et al.Mouth pressure in response to magnetic stimu-lation of the phrenic nerves. Thorax 1995;50:620-4.

76 Decramer M, Scano G. Assessment of respira-tory muscle function. Eur RespirJ 1994;7:1744-5.

77 Similowiski T, Duguet A, Straus C, Attali V,Boisteanu D, Derenne J-Ph. Assessment of vol-untary activation of the diaphragm usingcervical and cortical magnetic stimulation. EurRespirJ 1996;9:1224-31.

78 Laghi F, Harrison MJ, Tobin MJ. Comparisonofmagnetic and electrical phrenic stimulation inassessment of diaphragmatic contractility. JfAppl Physiol 1996;80: 1731-42.

79 Gea JG, Espadaler JM, Guiu R, Aran X, SeoaneL, Broquetas JM. Diaphragmatic activity in-duced by cortical stimulation: surface versusesophageal electrodes. J Appl Physiol 1993;74:655-65.

80 Oster G, Jaffe JS. Low frequency sounds fromsustained contraction of human skeletal mus-cles. BiophysJ7 1980;30:119-27.

81 Maton B, Petitjean M, Cnockaert JC. Phono-myogram and electromyogram relationshipswith isometric force reinvestigated in man. EurRespirJ 1990;60: 194-201.

82 Petitlean M, Bellemare F. Phonomyogram of thediaphragm during unilateral and bilateralphrenic nerve stimulation and changes withfatigue. Muscle Nerve 1994;17:1201-9.

on 17 July 2018 by guest. Protected by copyright.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.74.870.208 on 1 A

pril 1998. Dow

nloaded from