E f f e c t i v e p u l m o n a r y c a p i l l a r y blood flow i m p e d a n c e cardiogr:tphy

P e d i a t . Res. 11: 1171-1 177 (1977)

neonate nitrous oxide

Cardiac Output in the Neonatal Period Using Impedance Cardiography

S. N . MOHAPA.I'KA ANI> I ) . \V. HII.1. o f t < ~ ~ , q / o tl ~. iiti~, ~ ~ o~/ItI~I It,i(~/(/\, '\ ~.OII(/OII, t> the machine; the first deriv:rtive of the impedance carcliogriim. dZ/rlt o h m see I : a n EC< recorded h e t \ \ e c n thc o u t e r two electrodes and ;I phonoc:trdiogriiri~. the Iirttcr t \ \ o sigrlirls being usccl as timing devices. I t \\its irnpo\sihlc to put two electrodes a r o u n d the short neck of ;I young hahy. T h e first clectroclc \v:is therefore placed on the I>ro\\. I t \\;is found thirt the mean tr:rnsthoritcic impediince viilirc. Z,, o h m s . w;rs Illore \t;rhle. ;ind that ;I better qui~lityof r e u ) r J i n g \\;is ol~tiiincd if electrode jelly \\as applied t o the l v o \ \ clcctrode. I ' h c ;ipp;iratus \ + a s ;rl\v;iys sivitched o n irnd the c l e c t n ~ d c sa p p l i c ~ lfor :it lei~st 10 min twfore a recording was ~;~hc.n. Impedance cardiogrirrns \ \ e r e oht;iincd \\bile the child lay ~ l u i c t l y .any muscular irctivity causing gross distortion of the trace. A tvl>ic;~lsatisfactory trace obtained is s h o n n in Figure I . A s the hahy hrcirthed. ther-c \\:is a n inevitable swing of the hascline around the r e r o line and in this study care was taken only to itnalyzc I>c;its which \ \ e r e not distorted hy the respiratory s\\ing. I f this criterion was fulfilled, the bc;it \ \ a s analyzed. the height of d Z / d t rnin tlcing rncasured from the cornniencerncnt o f the s t e e p i ~ p s t r o k c .iclentificcl u i t h the help of the ECCi a n d 1,honocarcliogr;1111.~ r r c s p c c t i v cof the position of this point \\ith reslwct to the h ; ~ s e l i n c .'l'his is in contrast to the usual m e t h o d of an:il>zing adult i m p c ~ l a n c cciircliogranls n h i c h a r c Icss ;ifc height. d Z / d t rnin. is fected h! respiratory swings a n d ~ h e r the mc;ixurcd from thc I>asclinc. (';lrc i t a s taken to irnirly~c the

1171

previously described in Figure 2 and it can he seen that thesc data givc a higher v:~lucfor the resistivity than any of the prcvious d a t a . 'I'hc ICO's reported in this study itre denotctl in four different n a y s . u\ing four different resistivity value\ al>l>rcvi;~tctl thus: I C O ( K ) = I C O calculated using resistivit! data of Kuhicek ( 17); I C O (C;&S) = I C O calculated using resistivity d;~t;i of Getldcs and Sadler ( I I ) ; IC'O ( M & H ) = IC'O calculatetl using rczistivity data of Moh;rpatr;i and Hill ( 7 1 ) from worh with hlood of a n e ~ n i cdialysis patients: IC'O ( M . C & H ) = I C O calculated using resistivity data of Mohapiitr;~.C'osteloe. and Hill ( 2 2 ) from blootl samples. work with pl;icent;~l;11it1neo~i;it;~l h 1 t A S U K t t v l t N I O t Opc c l f

Fig. I . Typical impedance c;rrdiogram from a quretl> hrcathing h;~h!. showing n1c;isurcmcnt of the v;~luc\drldt min ;crid T.

The e\tirn;rtion of the effective pulmonary c:ipill;iry hlood flow. O p c eff, b) the rebreathing of nitrous oxide. N,O. relies upon the principle that the gas is ahsorhed into the hlooclstrcam in proportion to thc pulmonary hlood flo\v ant1 the concentration

same number of heat\ ;ihovc ant1 helo\\. the ha\cline \o t h ; ~ t variation of stroke volume during the respirator! cycle \ \ ; I \ :illo\ved for. It \vas very unusual for ;I tr;icc to he \o cli\tortetl 1 respiration that i t could not he iinalyzcd and u.;u:illy 8-1 7 bcats \\ere analyzed. \\,as c;~lculatcd For each beat analyzed :!ic stroke v o l u ~ n c AV. . from the crluation

where p is the resistivity of human hlootl at 37" in o h m ern. 1. is thc nleiin disti~nccbctweer~thc t\vo inner recording elcctrotlcs in centimeters. Z,, is the mean transthoracic impedance in o h m s . d Z / d t min is the maximum negative value of dZ/clt occurring during the cardiac cycle in o h m s sec I , and T is the ventricular ejection time in second\. The values of d Z / d t min a n d T \\ere obtained from the tracing as sholvn in Figure I and the I. S . E . 4 0 1 0 . arld tracings

> 31,:

800

.

400

-

< :Y

..

HCI

> 35%

EI

-

I r4

-

0

" -

D fl

100

400

bOO r n ~r n l n

'

100

0

i Oil

400

bOO nln

900

OPC EFF OPC EFF Fig. 3 . calculating the cocfficicnt of variation of paired cstirnatcs ( 8 ) (C'VPE) on the same data by the formula. C V P E = (SDd/X) x IOOCt . where SIId i4 the standard deviation of the difference het\\ccn the pairs of obscrvations and i\ thc 1ne311of all the observ;~tio~ls. The rc\ults of this analysis are given in T i ~ b l c1 which sho\\s good reproducibility for hoth mcthods of mcawring cardiac output.

The infants included in this study were all patient\ on the Premature Baby Unit at the Harnrncr\niith Hospital. There \\as

no selection in rcspcct of gcst;~tional age, birth weight. o r weight and age at the time of testing, and none hod clinical evidence of cardiovascul;tr or respiratory abnormality. Any baby in whom left t o right shunting was indicated by the recirculation of N,O carrying blood before ccluilihration had been ;~chieved,was excluded from the study. The techniilue of estimation of Qpc eff by rebreathing N,O has been used with healthy babies by other \vorkcrs. 'l'he use of the imped:~ncc method caused n o additional risk o r discomfort and both rnethods have been approved by the Ethics Committee of this hospital.

A total of 109 sirnult:~neous measurements of o p c eff and 1CO from 32 different infants studied upon 41 occasions were consigered for correlation purposes. T h e data for these infants and Qpc eff and I C O determinations using the different resistivity values arc given in T;~hle7 and the different methods are plotted against each other in the graphs (Figure 3 ) shoiving the linc of identity with f 1 0 % deviation. It is seen t h ; ~ t .using the Kubicck ( 1 7 ) and the Geddes and Sndler ( I I ) resistivity data. the impedance technique underestimates v;~luesin those infants whose hcmatocrit was below 3 5 % . The Mohapatra and Hill ( 2 1 ) data produces a greater tendcncv to undcrcstimation. In contrast. using the Mohapatra. Costcloc. and Hill ( 2 2 ) data from work with neonatal blood samples. the correlation is ;~cceptablein the anemic group but 3 5 % group. the I C O is overestimated in the Hct --, Using the mean value for Opt eff each time an infant was studied as thc standard. :I correction factor. S. to be applied t o the I C O calculated using the accurate resistivity data. p [ I C O (M.C&H)] (Fig. 2 ) . was derived. so t h t ~ tfor cach set of values. S = [Opt eff)/(lCO ( M . C & H ) ] . The values obtained for S for each occasion an infant studied were then related t o the relevit~lt hematocrit value and nonlinear rcgrcssion analysis to find the linc of hest fit was performed (Fig. 4 ) .

S = 2 62L n =Ll

,

-

1 121 Log

r =-0756

10

HL! % Ple esercise. T h e 1:trgc scatter of the points in the C'olton 1.1 trl. I97 1 ( 6 ) I3rndy and RRig;rtto. 1071 ( 3 ) d a t ~ from i which the constant. S. was derived (Fig. 1)is thought to reflect the nature ;tnd v a r i c t of the factors heing corrected. Orme 1.r (11.. 1073 (23) Wc n o w use this correction ktctor for all IC'O citlcul;itions. noting that its use is valid only if the irnpedaricc measurements c)f tllc ri\L\ o f ; ~ n c \ t h c - ;Ire made a \ i l l thi\ \lucly, I . , . .. 1))' mc;trul-ing the 111c;111~ l i \ l ; t r l ~ c c:~rcli;tcoutput c)n t1orrn;tl infant\ I~cc:ru\c~ sia and catheterization; thus the noni11v;tsive technielue of cstih c t ~ e c nthe rccortling electrodes. estim:tting the hcm;ttocrit o n mation of Qpc eff was used its n standard i n this sturly. I t is a specimen of blood obtained by prick using a lancet rather gcneritlly ackno\vledged that this method clocs not give it v;ilicl than hy verlipuncturc, and 31sing the relationship bct\vccn resisindication of cardiac output when there is respiratory o r cardiotivity of blood and hematocrit. p = 0 7 . 0 1 9 c x p ( 0 . 0 2 4 3 3 H c t ) vascular abnormality, hut in norni;il adults i t has heen shown to derived from \\ark \\ith fresh neonatal hlood samples ( 2 2 ) . correlate well with invasivc measurements of cardiac output Impedance c:trdiogr:iphy and rchrcathing techniclucs arc prefboth :it rest ( 2 0 ) and with cscrcise ( 2 ) . erahlc to other methods of estimating c:trdiac output in the The technique as described in this stucly \\.;is highly sensitive nc\\horn in that they arc noninvasive. 1rnpcd:ince cardiogriiphy in detecting any persistent left to right fetirl shunt. which is easier to perform than ;I rchreiithing proccdure and once the involved exclusion from the study. This complic:ition was enelectrodes arc applierl, rcpc:ited measurements can he m;tdc countered in 2 0 r i of prctcrm infants. up to 2 weeks of postniital uithout any further hitndling of the hiihy. In addition. i t hiis the age. hut heyond this age i t \\as rare. T h e IC'O in these h;thics irdvant:t_ce over all o t h c r method5 of providing cluantitiitivc heat tended to he highcr than in those who :vcrc not shunting. 'l'hia to heat information as opposerl to a mean tlon. In this age phenomenon reduces the potential usefulness of the irnpcilance group i t is essential to cstimirte the hematocrit hecause of the cardiograph for quantitatively accurate measurcnicnt of cardiac wide range encountered. hut even if accurate value\ for resistivoutput in vcr) young preterm hithics. 'l'he results of ( j p c cff it! arc used. a further hcmatocrit-related correction factor estimation were highly reproducihle and compare fitvorahly should he applied for optimal results. n i t h similar measurerncnts made hy other \vorkcrs ( 3 . 1. 6. 2 3 ) . T h e rnethod \ \ a s therefore regarclcd as hcing ir sirtisfactory standard. T h e subdivision of the infants into t\\o groups o n the 1);tsis of I . H;lhcr. I . t.,Judy. W . V . . (iedde\. 1.. I:.. Langlcy. I-. hl., and Hill. 1). W . : hematocrit \\;is done hecituse of the revealed inaccuritcics of the hle;l\urement o f cardlac output h) mean\ of clcctric:~l impedance. ('ard ~ o v : ~ rR. e \ . Center Hull. V : I35 (1971 ). impedance method which appearetl to he hematocrit related. 2 . Hecklake. M . R.. J i t r b i s . (' I.. Pengelly. l..I).. Kclining, S . . Mccircgor. M . . The t\4o groups \\ere compared in other aspects. There h a s n o ;lnd B ; I I ~ \ .D . V . : Me;~\urement o f pulrnon:~ry hlood flou during excrcrse significant difference between the weight and length at testing. wing nltrous oxide J Appl. Phy\iol:, 17: 57Y.( l Oh?). I' 0 . 1 . T h e infants in the lo\v hem;ttocrit group were less 3 . Ur:~dy. J . P . , 'tnd Rigatto. t i . : Pulmonary c:~pillary hlc~odflew i n newhorn ~ n f a n t uslnp s plethy\mogr;~phy; ~ n dnltrous oxide. Pediatr~cs.4 8 . 207 ( 1071 ). mature. lighter at birth. and older at the time of testing, I' < 4. C h u . J . . Clement\. J . A . . ('lcll) using :I stand:trcl tcchnicluc ( 2 ) . 'The breast \v:t> c1c;ine~l ~ i t h water only a n d 5-ml samples were taken into 4tcrilc universal containers i i p p r o s i m ~ r t e l 1-2 ~ min after starting the fccd ; ~ n d; I t