ISSN 0003-3022
`
`January 1989
`ANESAV
`
`
`
`The Journal of
`The American Society of
`Anesthesiologists,
`Inc.
`
`
`
`Volume 70
`
`°
`
`Number1
`
`Complete Contents
`
`Pages 3, 5, 7,9, and 11
`
`1
`
`APPLE 1011
`
`1
`
`APPLE 1011
`
`

`

`Anesthesiology
`
`THE JOURNAL OF
`THE AMERICAN SOCIETY OF ANESTHESIOLOGISTS,INC.
`Editor-in-Chief
`LAWRENCEJ. SAIDMAN, M.D., San Diego, California
`Editors
`David E. Longnecker, M.D.
`Julien F. Biebuyck, M.B., D.Phil.
`Philadelphia, Pennsylvania
`Hershey, Pennsylvania
`Dennis T. Mangano, Ph.D., M.D.
`John J. Downes, M.D.
`San Francisco, California
`Philadelphia, Pennsylvania
`Kai Rehder, M.D.
`H. Barrie Fairley, M.B., B.S.
`Rochester, Minnesota
`Stanford, California
`Donald R. Stanski, M.D.
`Simon Gelman, M.D., Ph.D.
`Stanford, California
`Birmingham, Alabama
`Michael M. Todd, M.D.
`Carol A. Hirshman, M.D.
`Iowa City, lowa
`Baltimore, Maryland
`Warren M. Zapol, M.D., Boston, Massachusetts
`
`
`
`and
`
`i
`
`Associate Editors
`Charles W. Buffington, M.D.
`Carl Lynch III, M.D., Ph.D.
`Pittsburgh, Pennsylvania
`Charlottesville, Virginia
`David H. Chestnut, M.D.
`Mervyn Maze, M.B., Ch.B.
`Stanford, California
`lowa City, lowa
`Henry Rosenberg, M.D.
`Jeffrey B. Cooper; Ph.D.
`Philadelphia, Pennsylvania
`Boston, Massachusetts
`Gary R.Strichartz, Ph.D.
`Dennis M.Fisher, M.D.
`Boston, Massachusetts
`San Francisco, California
`Tony L. Yaksh, Ph.D.
`Thomas F. Hornbein, M.D.
`San Diego, California
`Seattle, Washington
`
` correspondencerelating to editorial management,and Letters to the Editor should be mailed to Lawrence
`Manuscripts for publication,
`T-015, University of California—San Diego, San Diego, California 92093. Books for review should
`J. Saidman, M.D., Anesthesiology,
`1
`The Children’s Hospital of Philadelphia, One Children's Plaza, 34th
`be mailed to John J. Downes, M.D., Department of Anesthesia,
`Street and Civic Center Boulevard, Philadelphia, PA 19104.
`—
`All articles are accepted for publication with the understanding that they are contributed exclusively to this Journal and become the
`property of the American Society of Anesthesiologists, Inc. Statements or opinions expressed in the Journal reflect the views of the
`author(s) and do not represent official policy of the American Society of Anesthesiologists unless so stated.
`For details and recommendations concerning the preparation of manuscripts see GUIDE FOR AUTHORS, which appearsin the January
`ehecuanasde related correspondence should be addressed to Advertising Manager, ANESTHESIOLOGY,J. B. Lippincott Company,
`East Washington Square, Philadelphia, Pa. 19105. Publication ofan advertisementin ANESTHESIOLOGYdoesnot constitute endorsement
`by ANESTHESIOLOGYofthe product described therein or of any representations madeby the advertiser with respect to the product.
`(003-3022) is published 13 times a year (monthly except September, in which two issues are published)
`esthesiologists, Inc., by J. B. Lippincott Company, at Downsville Pike, Route 3, Box 20-B, Hagerstown,
`ANESTHESIOLOGY (ISSN 0
`East Washington Square, Philadelphia, PA 19105. Printed in the U. S. A. © Copyright 1989
`for the American Society of An
`MD 21740. Business offices are located at
`s, Inc. Secondclass postage paid at Hagerstown, MD,andat additional mailing offices.
`by the American Society of Anesthesiologist
`ndence related to membership and change of address should be mailed to the Executive
`Notice of Change of Address: All correspo!
`of Anesthesiologists, 515 Busse Highway, Park Ridge, Illinois 60068. Correspondence regarding nonmember
`S
`j
`very
`aibecrieersact agen iJournal Fulfillment Department, Lippincott/Harper, Downsville Pike, Route 3, Box 20-B, Hagerstown,
`Maryland 21740,or call 1-800-638-3030; in Maryland, call collect 301-824-7300. Correspondence regarding nonmembersubscriptions
`inJapan,including inquiriesaboutsubscriptionratesandsubscription orders,should besentto WoodbellIncorporated, 4-22-11, Kitakasai,
`Edogawaku, Tokyo 134, Japan. In India, Nepal,Eanes and Sri Lanka contact Universal Subscription Agency Pvt. Ltd., 101-102
`C
`as
`ew Delhi-110017,
`India.
`Iemmane—aOw individual, $50.00institution; all other countries exceptJapan, India, Nepal, Bangladesh, and
`Sri Lanka, $75.00 individual, $85.00 institution. Single copies $8.00. Rates for air mail delivery available upon request.
`—
`Ifthe Society Office or the Publisher receives a request for replacement within 60 days ofthe mailing in the U.S. or within 90 days in
`i
`iginal
`i
`is returned by
`the
`postal service as undeliverable, the issue will be replaced.
`Sonne aaae ccrorhestology: downsville Pike, Route 3, Box 20-B, Hagerstown, MD 21740.
`Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted byThe
`American Society ofAnesthesiologists for librariesand other users registered with the CopyrightClearance Center (CCC) Transactional
`Reporting Service, provided that the base fee of $00.80 per copy, plus $00.05 per page1s paid directly to CCC, 21 CongressSty Salem,
`MA 01970, 0003-3022/89 $00.80 + $00.05. This does not extend to other kinds ofcopying, such as copying for general distribution,
`for advertising or promotional purposes, for creating new collective works, or for resale.
`
`
`
`2
`
`2
`
`

`

`Anesthesiology
`
`January 1989
`
`
`
`CONTENTS
`
`EDITORIAL VIEWS
`
`A Changein Formatfor ANESTHESIOLOGY.
`LawrenceJ. Saidman
`Changing Perspectives in Monitoring Oxygenation.
`
`H. Barrie Fairley
`Studies in Animals Should Precede Human UseofSpinally Administered Drugs.
`Tony L. Yaksh andJ. G. Collins
`
`9
`
`4
`
`7
`
`13
`
`19
`
`CLINICAL INVESTIGATIONS
`The Influence of Renal Failure on the Pharmacokinetics and Duration of Action of
`Pipecuronium Bromide in Patients Anesthetized with Halothane and Nitrous
`Oxide.
`James E. Caldwell, P. Claver Canfell, Kay P. Castagnoli, Daniel P. Lynam, Mark R.
`Fahey, Dennis M.Fisher, and Ronald D. Miller
`Effect of Intercostal Nerve Blockade on Respiratory Mechanics and CO, Chemosen-
`sitivity at Rest and Exercise.
`Bernice R. Hecker, Robert Bjurstrom, and Robert B. Schoene
`ocardiographic, Mechanical, and Metabolic In-
`Comparison of Hemodynamic, Electr
`dicators of Intraoperative Myocardial Ischemia in Vascular Surgical Patients with
`Coronary Artery Disease.
`Soren Haggmark, Per Hohner, Margareta Ostman, Arnold Friedman, George Diamond,
`Edward Lowenstein, and Sebastian Reiz
`mall Doses of Sufentanil or Fentanyl: Dose Versus EEG
`Induction of Anesthesia with S
`and Thiopental Requirement.
`Response, Speed of Onset,
`T. Andrew Bowdle and RichardJ. Ward
`Relationship of Mivacurium Chloride in Humans during Nitrous
`Oxide-Fentany!or
`Nitrous Oxide-Enflurane Anesthesia.
`Mark R. Fahey, Daniel P. Lynam, and
`James E. Caldwell, John B. Kitts, Tom Heier,
`Ronald D. Miller
`Transurethral Resection of the Prostate, Serum Glycine Levels, and Ocular Evoked
`Potentials.
`Janice Mei-Li Wang, DonnellJ. Creel, and K. C. Wong
`(Continued onpage 5)
`———ee
`
`The Dose-Response
`
`26
`
`31
`
`36
`
`3
`
`

`

`January 1989ANESTHESIOLOGY
`
`CONTENTS
`
`(Continued from page 5)
`
`Lower Esophageal Contractility Predicts Movement during Skin Incision in Patients
`Anesthetized with Halothane, but Not with Nitrous Oxide and Alfentanil.
`Daniel I. Sessler, Randi Sten, ChristineI. Olofsson, and Franklin Chow
`
`Determination of Intra-abdominal Pressure Using a Transurethral Bladder Catheter:
`Clinical Validation of the Technique.
`ThomasJ. Iberti, Charles E. Lieber, and Ernest Benjamin
`
`LABORATORY INVESTIGATIONS
`Epidural Clonidine Analgesia in Obstetrics: Sheep Studies.
`James C. Eisenach, Maria I. Castro, David M. Dewan, andJames C. Rose
`The Enhancementof Proton/Hydroxyl Flow across Lipid Vesicles by Inhalation An-
`esthetics.
`Douglas E. Raines and David S. Cafiso
`The Influence of Dextrose Administration on Neurologic Outcome after Temporary
`Spinal Cord Ischemia in the Rabbit.
`John C. Drummond and Suzanne S. Moore
`Tachyphylaxis to Local Anesthetics Does Not Result from Reduced DrugEffectiveness
`at the NerveItself.
`Peter Lipfert, Holger Holthusen, andJoachim O. Arndt
`Comparisonof the Effects of Halothane on Skinned Myocardial Fibers from Newborn
`and Adult Rabbit. L. Effects on Contractile Proteins.
`ElliotJ. Krane and Judy ¥. Su
`Regional Differences in Left Ventricular Wall Motionin the Anesthetized Dog.
`Johan Diedericks, BruceJ. Leone, and Pierre Foex
`e’”’ on Nitrous Oxide Anesthesia, Tolerance, and Physical De-
`Effects of ‘‘Nitrendipin
`pendence.
`S.J. Dolin and H. J. Little
`
`MEDICAL INTELLIGENCE ARTICLE
`
`Pulse Oximetry.
`Kevin K. Tremper and Steven J. Barker
`
`42
`
`47
`
`5]
`
`-
`
`“a
`
`we
`
`"6
`
`82
`
`af
`
`98
`
`i —
`
`(Continued on page7)
`
`4
`
`

`

`January 1989 ANESTHESIOLOGY
`
`CONTENTS
`
`(Continued from page 5)
`
`LABORATORY REPORTS
`Laudanosine Does Not Displace Receptor-specific Ligands from the Benzodiazepinergic
`or Muscarinic Receptors.
`Yeshayahu Katz and Moshe Gavish
`Effects of Methemoglobinemia on Pulse Oximetry and Mixed Venous Oximetry.
`Steven J. Barker, Kevin K. Tremper, and John Hyatt
`Hyperbilirubinemia Does NotInterfere with Hemoglobin Saturation Measured by Pulse
`Oximetry.
`Francis Veychemans, Philippe Baele, J. E. Guillaume, Eric Willems, Annie Robert, and
`Thierry Clerbaux
`Evaluation of a Blood Gas and Chemistry Monitor for Use during Surgery.
`G. Bashein, Wesley K. Greydanus, and Margaret A. Kenny
`A Model for Determining the Influence of Hepatic Uptake of Nondepolarizing Muscle
`Relaxants in the Pig.
`Johann Motsch, PimJ. Hennis, Franz Alto Zimmermann, and Sandor Agoston
`Transtracheal Doppler: A New Procedure for Continuous Cardiac Output Measure-
`ment.
`Jerome A. Abrams, Roland E. Weber, and Kenneth D. Holmen
`
`CASE REPORTS
`
`Managementof Acute Elevation of Intracranial Pressure during Hepatic Transplan-
`tation.
`D. Brajtbord, R. 1. Parks, M. A. Ramsay, A. W. Paulsen, T. R. Valek, T. H. Swygert,
`and G. B. Klintmalm
`Treatmentof Isorhythmic A-V Dissociation during General Anesthesia with Propran-
`olol.
`Russell F. Hill
`
`00
`
`112
`
`a6
`
`123
`
`8
`
`ae
`
`139
`
`14]
`
`144
`
`146
`
`(Continued onpage 9)
`
`Fiberoptic Endobronchial Intubation for Resection ofan Anterior Mediastinal Mass.
`Dirk Younker, Randall Clark, and Lewis Coveler
`Postpartum Seizure after Epidural Blood Patch andIntravenousCaffeine Sodium Ben-
`on, Craig H. Leicht, and Thomas S. Scanlon
`
`zoate.
`
`Vincent E. Bol
`
`it
`
`5
`
`

`

`January 1989ANESTHESIOLOGY
`
`CONTENTS
`
`.
`7
`Caudal Epidural Anesthesia in an Infant with Epidermolysis Bullosa.
`Lawrence L. Yee, Joel B. Gunter, and Charles B. Manley
`
`(Continued frompage 7)
`ia
`
`Recurrent Respiratory Depression after Alfentanil Administration.
`Rory S. Jaffe and Dennis Coalson
`Pain of Delayed Traumatic Splenic Rupture Maskedby Intrapleural Lidocaine.
`William W. Pond, Gregory M. Somerville, Siong H. Thong, James A. Ranochak, and
`Gregory A. Weaver
`Dose-response Relationship for Succinylcholinein a Patient with Genetically Determined
`Low Plasma Cholinesterase Activity.
`Charles E. Smith, Geraint Lewis, Francois Donati, and David R. Bevan
`
`CORRESPONDENCE
`Determination of Decay Constants from Time-varying Pressure Data.
`Charles Beattie, Linda S. Humphrey, and Gary Maruschak
`Reply. Clifford R. Swanson and William W. MuirIIT
`Use Caution when Extrapolating from a Small Sample Size to the General Population.
`David J. Benefiel, Edward A. Eisler, and Rodger Shepherd
`Reply. Imad H. Abdul-Rasool, Daniel H.Sears, and Ronald L. Katz
`Succinylcholine and Trismus.
`Frederic A. Berry and Carl Lynch III
`Reply. Henry Rosenberg
`An Infant Modelto Facilitate Endotracheal TubeFixationin the Pediatric ICU Patient.
`Patrick K. Birmingham and Babette Horn
`An Alternative Method for ManagementofAccidental Dural Puncture for Labor and
`Delivery.
`Shaul Cohen, Jonathan S. Daitch, and Paul L. Goldiner
`High-pressure Uterine Displacement.
`Michael J. Dorsey and Walter L. Millar
`Calculating the Potency of Mivacurium.
`Aaron F. Kopman
`Reply. John J. Savarese
`
`151
`
`154
`
`Ke
`
`159
`
`166
`160
`
`161
`a
`o
`163
`
`‘ee
`
`16
`
`166
`
`166
`
`(Continued on page 11)
`i
`
`6
`
`

`

`January 1989—ANESTHESIOLOGY
`
`CONTENTS
`
`|
`
`|
`
`Midazolam in a Malignant Hyperthermia-susceptible Patient.
`Juliana H. J. Brooks
`Exchange Autotransfusion Using the Cell Saver during Liver Transplantation.
`Marc R. Brown, Michael A. E. Ramsay, and Thomas H. Swygert
`Air Entrainment Through a Multiport Injection System.
`Dean Gilbert, TheodoreJ. Sanford, Jr., and Brian L. Partridge
`Reply. Thelma Macedo
`A Tracheal Tube Extension for Emergency Tracheal Reanastomosis.
`
`Robert S. Holzman
`
`The Relationship Between Malignant Hyperthemia and Neuroleptic Malignant Syn-
`drome.
`Haggai Hermesh, Dov Aizenberg, Margo Lapidot, and Hanan Munitz
`Reply. Stanley N. Caroff, Stephan C. Mann, Henry Rosenberg, Jeffrey E. Fletcher, and Terry
`D. Heiman-Patterson
`
`Appropriate Facilitation of Intravenous Regional Techniques in RSD.
`Kevin Foley, Linda Schatz, and Randall L. Martin
`
`REPORTOF SCIENTIFIC MEETING
`
`ANNOUNCEMENT
`
`1]
`
`(Continued from page 9)
`
`167
`
`168
`
`169
`170
`170
`
`_
`
`(ep
`
`173
`
`174
`
`173
`
`issue,
`
`The Guide for Authors is published in the January andJuly issues. It may be found on page 33A ofthis
`
`GUIDE TO AUTHORS
`
`ANESAV is a code word (“coden") used by the Chemical Abstract Service to identify the journal.
`
`ae
`
`7
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`8
`
`

`

`Anesthesiology
`V 70, No 1, Jan 1989
`
`PULSE OXIMETRY
`
`99
`
`
`
`
`
`
`%O,SATURATION
`
`
`
`
`ny|AE1 :
`
`
` oeEe
`
`
`
`i ee
`
`
`
`leoes
`ate
`hal
`
`
`
`
`
`
`
`
`-TCoe H gees
`
`
`
`
`WO ay 16
`i)
`
`me
`
`12
`
`iron
`
`Fic. 1. This figure is from a 1951 article in ANESTHESIOLOGY. It
`reveals dramatic desaturation in a 4-yr-old patient during a tonsillec-
`tomy. Reproduced from Stephen RC, Slater HM, Johnson AL,Sekelj
`P; The oximeter—A technicalaid for the anesthesiologist. ANESTHE-
`SIOLOGY 12:548, 1951, with permission.
`
`The Physics and Physiology of Pulse Oximetry
`
`BEER’s LAw
`
`In the 1930s, Matthes used spectrophotometry to de-
`termine hemoglobin oxygen saturation.? This methodis
`based on the Beer-Lambert law, which relates the con-
`centration of a solute to theintensity oflight transmitted
`through a solution.
`
`Tans = ie
`
`(1)
`
`(la)
`A = DCe
`whereI,rans = intensity of transmittedlight;Ij, = intensity
`of incidentlight; A = absorption; D = distancelightis
`transmitted through the liquid (path length); C = con-
`centration of solute (hemoglobin); ¢ = extinction coeffi-
`cient ofthe solute (a constantfor a given solute at a spec-
`ified wavelength). Thus, if a knownsoluteis in a clear
`solution in a cuvette of known dimensions, the solute con-
`centration can be calculated from measurements of the
`incident and transmitted light intensity at a known wave-
`length, The extinction coefficient¢ is a property oflight
`absorption for a specific substance at a specified wave-
`length. In a one-component system, the absorption A is
`the product of the path length, the concentration, and
`the extinction coefficient, equation la. If multiple solutes
`
`becamea standard clinical and laboratory tool in pul-
`monary medicine. Although it was demonstrated to be
`accurate for intraoperative monitoring,’® its size and ex-
`pense, and the cumbersomenature of the ear probe pre-
`vented its acceptance as a routine monitor. At this time,
`all oximeters produced variouslight source wavelengths
`by filtering white light. The filtered light was then trans-
`mitted to and from thetissue throughfiberoptic cables.
`In the mid 1970s, Takuo Aoyagi, an engineer working
`for Nihon Kohden Corporation, made an ingeniousdis-
`covery regarding oximetry. He was developing a method
`to estimate cardiac output semi-noninvasively by detecting
`the washout curve of dye injected into a peripheral vein
`as it perfused the ear. This washout curve was measured
`in the ear with a red andinfrared light densitometer sim-
`ilar to the Millikan ear oximeter. He noticed that his
`washoutcurves contained pulsations dueto thearterial
`pulse in the ear. To moreeasily analyze the dye washout
`curve, he subtracted these pulsations fromthe curve, and
`in doing so he discovered that the absorbanceratio of the
`pulsations at the two wavelengths changed with arterial
`hemoglobin saturation. He soon realized that he could
`build an ear oximeter that measuredarterial hemoglobin
`saturation withoutheating the ear by analyzingpulsatile
`light absorbances.” This first pulse oximeter, developed
`by Nihon Kohden, used filtered light sourcessimilar to
`Millikan’s ear oximeter. The device was evaluated clini-
`cally in the mid 1970s and marketed withlittle success.”
`In the late 1970s, Scott Wilber in Boulder, Colorado,
`developedthefirst clinically accepted pulse oximeter by
`making two modifications of
`the Nihon Kohden
`method.''+ First, he produced a lightweight sensor by
`using light emitting diodes (LEDs) as light sources and
`photodiodesas detectors. Consequently, the instrument
`was connectedtoits earclip sensor only by a smallelectrical
`cable. Wilber also improved thesaturation estimates by
`using a digital microprocessorto store a complex calibra-
`tion algorithm based on humanvolunteer data.'* This
`methodwill be discussed in more detail below. This device
`was developedby Biox Corporation of Boulder, Colorado,
`and wassuccessfully marketed to pulmonary function lab-
`oratories in the early 1980s.
`Theclinical utility of the noninvasive oximeter in the
`operating room wasrediscoveredin the 1980s by William
`New,an anesthesiologist at Stanford University. Realizing
`that a continuous, noninvasive monitor of oxygenation
`would be useful to anesthesiologists, New developed and
`marketed a pulse oximeterto this group.'® The Nellcor
`model N100 had by 1985 become almost synonymous
`with the term “‘pulse oximeter.”
`
`+ Wilber S: Blood constituent measuring device and method. US
`Patent #4, 407, 290 April 1, 1981.
`
`9
`
`

`

`K. K. TREMPER ANDS. J. BARKER
`
`Anesthesiola,
`V 70, No 1, Jan 1989
`
`HEMOGLOBIN EXTINCTION CURVES
`
` hemoglobin ExtinctionCoefficient
`
`
`
`
`04
`600 640
`
`:
`
`carboxyhemoglobin |
`—.
`z
`
`680
`
`720
`
`760
`
`800
`
`840
`
`880
`
`920
`
`960
`
`By the above definition of oxygen saturation, the two
`forms of hemoglobin that do not bind oxygen (COHb
`and MetHb)are notincluded. This is the origin of what
`is now referred to as “functional hemoglobin saturation,”
`defined as (Severinghaus JW, personal communication);
`
`With the advent of multiwavelength oximeters that can
`measureall four species of hemoglobin, “fractionalsat-
`uration” has been defined as the ratio of oxyhemoglobin
`to total hemoglobin:
`
`Infaed
`
`methemoglobin
`
`|| |
`
`| |
`
`
`
`
`
`
`OsHb
`oxyhemoglobin
`onto e eee ane,
`Functional SaO, =O.Hb+HbxX 100%.
`(2)
`
`reduced
`!
`\
`
`Wavelength \ (nm)
`
`Fractional SaO»
`
`Fic, 2. Transmitted light absorbance spectra of four hemoglobin
`species: oxyhemoglobin, reduced hemoglobin, carboxyhemoglobin, and
`methemoglobin. Adapted from Barker SJ and Tremper KK: Pulse
`Oximetry: Applications and limitations, Advances in Oxygen Moni-
`toring, International Anesthesiology Clinics. Boston, Little, Brown and
`Company, 1987, pp. 155-175.
`
`OsHb
`——————TT————OOoooTo * 100%.
`fs
`O2Hb + Hb + COHb + MetHb
`The fractional hemoglobin saturation is also called the
`“oxyhemoglobin fraction,” or ““oxyhemoglobin %.”’"*
`Whenoximetry is used to measure hemoglobin satu-
`ration, Beer’s law mustbe applied to a solution containing
`four unknownspecies: O2Hb, Hb, COHb, and MetHb.
`Expanding equation 1a toa four-componentsystem results
`in an absorption given by:
`
`A= D,Cye; ah DoCo€s + DsCseé5 + DsCaeq.
`
`(1b)
`
`The subscripts 1 through 4 correspond to the four he-
`moglobin species. If the path lengths are the same, then
`D can be factored out:
`
`A= D(Cye; + Co€9 Fi Cs€5 + C4€,).
`
`(1c)
`
`are present, A is the sum of similar expressions for each
`solute. The extinction coefficient can vary dramatically
`with the wavelength of the light. The extinction coeff-
`cients for various hemoglobin species in the red and in-
`frared wavelength range are shownin figure 2.
`Laboratory oximeters use this principle to determine
`hemoglobin concentration by measuring the intensity of
`light transmitted through a cuvette filled with a hemo-
`globin solution produced from lysed red blood cells."*
`For Beer’s law to be valid, both the solvent and the cuvette
`through e€4 are constantsat
`The extinction coefficients €,
`must be transparentat the wavelength used,thelight path
`a given wavelength2(fig. 2). The absorption defined in
`length must be known exactly, and no absorbing species
`equation lc is determined from equation 1 by measuring
`can bepresentin the solution other than the knownsolute.
`the incident and transmittedlight intensities. From equa-
`It is difficult to fulfill these requirementsin clinical devices;
`tion lc, we see that four wavelengthsoflight are needed
`therefore, each instrumenttheoretically based on Beer’s
`to produce four equationsto solve for the unknown con-
`law also requires empirical corrections to improve accu-
`centrations C,; through C4. If COHb and MetHb were
`racy.
`not present, their contributions to the absorption would
`be zero and functional hemoglobin saturation could be
`determined by a two-wavelength oximeter(two equations
`and two unknowns). If two wavelengths existed for which
`the extinction coefficients for COHb and MetHb were
`zero, then these absorption terms would again be zero
`anda two-wavelength oximeter could measure functional
`saturation. Unfortunately,as: illustrated in figure 2, the
`extinction coefficients for COHb and MetHbarenotzero
`in the red and infrared range, and their presencewill,
`therefore, contribute to the absorption. Even though the
`definition of functional hemoglobin saturation involves
`only two hemoglobin species (O2Hb and Hb), when
`.MetHb and COHbarepresent, four wavelengths are re-
`quired to determine either functional or fractional he-
`moglobinsaturation.'*
`
`HEMOGLOBIN SATURATION DEFINITIONS
`
`Adult blood usually contains four species of hemoglo-
`bin: oxyhemoglobin (OgHb), reduced hemoglobin (Hb),
`methemoglobin (MetHb), and carboxyhemoglobin (CO-
`Hb)(fig. 2). The last two species are in small concentra-
`tions, except in pathologic conditions. There are several
`definitions of hemoglobin saturation. Historically, “‘oxy-
`gen saturation” wasfirst defined as the oxygen content
`expressed as a percentage of the oxygen capacity. The
`oxygen content(cc of oxygen per 100 cc of blood) was
`measured volumetrically by the method of Van Slyke and
`Neill (1924).'° The oxygen capacity was defined as the
`oxygen content after the blood sample had been equili-
`brated with room air (158 mmHgoxygenatsealevel).
`
`10
`
`10
`
`

`

`Anesthesiology
`V 70, No 1, Jan 1989
`
`PULSE OXIMETRY
`
`101
`
`Absorption due to tissue
`
`PANN
`OYwyRON
`K?RRR
`
`Absorption due to pulsatile arterial blood
`Absorption due to non-pulsatile arterial blood
`%o%
`¢ | Absorption due to venous and capillary blood
`5
`ee
`/ AAA, rAt
`
`Time
`
`)"X
`
`LightAbsorption
`
`bc
`
`Fic. 3. This figure schematically illustrates the light absorption
`through living tissue. Note that the AC signalis due to the pulsatile
`componentofthe arterial blood while the DC signal is comprised of
`all the nonpulsatile absorbersin the tissue; nonpulsatile arterial blood,
`venousandcapillary blood,andall othertissues. Adapted from Ohmeda
`Pulse Oximeter Model 3700 Service Manual, 1986, p. 22.
`
`should be maximized.Asweseein figure 2, at 660 nano-
`meters, reduced hemoglobin absorbs about ten times as
`muchlight as oxyhemoglobin. (Note that the extinction
`coefficients are plotted on a logarithmicaxis.) At the in-
`frared wavelength of 940 nanometers, the absorption
`coefficient of OgHbis greater than that of Hb.
`
`Engineering Design and Physiologic Limitations
`__ Althoughthe theory on whichpulse oximetryis based
`is relatively straightforward, the application of this theory
`to the production ofa clinically useful device involves a
`
`
`
`SpOz(%)
`
`100
`90
`80
`70
`60
`
`50
`40
`30
`
`20
`10
`
`0
`
`0.4 0.6 0.8 1.01.2 1.4 1.6 1.8 2.0 2.22.4 2.62.8 3.0 3.2 3.4
`
`AC,,,,/DC 660
`AC,,,/DC 940
`
`Fic. 4. This is a typical pulse oximetercalibration curve. Note that
`the SaQgestimate is determined fromtheratio (R) of the pulse-added
`red absorbance at 660 nanometersto pulse-added infrared absorbance
`at 940 nanometers. The ratios of red to infrared absorbances vary
`from approximately .4 at 100% saturation to 3,4 at 0% saturation.
`Note thattheratio of red to infrared absorbanceis one at a saturation
`of approximately 85%. This curve can be approximately determined
`on a theoretical basis but, for accurate predictions of SpOg, experi-
`mental data are required. Adapted from JA Pologe: Pulse oximetry:
`Technical aspects of machine design, International Anesthesiology
`Clinics, Advances in Oxygen Monitoring. Edited by Tremper KK,
`Barker SJ. Boston, Little, Brown and Company, 1987, p 142.
`
`11
`
`PULSE OXIMETRY
`
`Noninvasive oximeters measurered andinfraredlight
`transmitted througha tissue bed,effectively using thefin-
`ger or ear as a cuvette containing hemoglobin. There are
`several technical problemsin accurately estimating SaO,
`by this method. First, there are many absorbers in the
`light path other than arterial hemoglobin, includingskin,
`soft tissue, and venousand capillary blood. Theearly ox-
`imeters subtracted the tissue absorbance by compressing
`the tissue during calibration to eliminate all the blood,
`and using the absorbance of bloodless tissue as the base-
`line. These oximeters also heated the tissue to obtain a
`signal related to arterial blood with minimum influence
`of venous and capillary blood.
`Pulse oximeters deal with the effects of tissue and ve-
`nous blood absorbances in a completely different way.
`Figure 3 schematically illustrates the series of absorbers
`in a living tissue sample. At the top ofthe figure is the
`pulsatile or AC component, whichis attributed to the
`pulsating arterial blood. The baseline or DC component
`represents the absorbances of the tissue bed, including
`venous blood, capillary blood, and nonpulsatile arterial
`blood. Thepulsatile expansion of the arteriolar bed pro-
`ducesan increase in pathlength (see equation 1b), thereby
`increasing the absorbance.All pulse oximeters assumethat
`the only pulsatile absorbance between the light source
`and the photodetectoris that of arterial blood. They use
`two wavelengthsoflight: 660 nanometers (red) and 940
`nanometers(near infrared). The pulse oximeterfirst de-
`termines the AC componentof absorbanceat each wave-
`length and divides this by the corresponding DC com-
`ponentto obtain a “‘pulse-added” absorbancethatis in-
`dependentofthe incidentlight intensity. It thencalculates
`the ratio (R) of these pulse-added absorbances, whichis
`empirically related to SaOg:
`
`—_ AC60/DCe6o
`za
`ACg40/DCo40
`
`.
`
`(4)
`
`Figure 4 is an example of a pulse oximetercalibration
`curve,'® The actual curves used in commercial devices
`are based on experimental studies in humanvolunteers.
`Note that when the ratio of red to infrared absorbance
`is one, the saturation is approximately 85%. This fact has
`clinical implications to be discussedlater.
`It is a fortuitous coincidence of technology and phys-
`iology that allowed the developmentofsolid-state pulse
`oximeter sensors.'® Light emitting diodes (LEDs) are
`available over a relatively narrow range ofthe electro-
`magnetic spectrum. Amongtheavailable wavelengths are
`somethatnotonly pass throughskin butalso are absorbed
`by both oxyhemoglobin and reduced hemoglobin. For
`best sensitivity, the difference betweenthe ratios of the
`absorbances of OgHb and Hb at the two wavelengths
`
`11
`
`

`

`102
`
`K. K. TREMPER AND S. J]. BARKER
`
`va eenE
`
`DYSHEMOGLOBINS AND DYES
`
`significant engineering effort. This section will present in
`general terms the clinical and physiologic problems of
`oximeter design and their engineering solutions. The dis-
`cussion is divided into four areas: dyshemoglobins and
`dyes, LED center wavelength variability, signal artifact
`management, and accuracy and response. The reader
`should be aware that these problems can interact with
`one another.
`
`4, an absorbanceratio of one correspondsto a saturation
`of 85% on the calibration curve. Pulse oximeter error
`during methemoglobinemiahas also been confirmedin a
`clinical report.'?
`In neonatalblood,a fifth type of hemoglobinis present,
`fetal hemoglobin (HbF). HbF differs from adult Hb in
`the amino acid sequences of two of the four globin sub-
`units. Adult Hb has two a- and two §-globin chains, while
`HbFhas two a and two f chains. This difference in globin
`chains haslittle effect on the extinction curves and there-
`fore shouldnotaffect pulse oximeter readings.§{ Thisis
`Being two-wavelength devices, pulse oximeters can deal
`indeed fortunate because the fraction of HbF presentin
`with only two hemoglobin species. As noted above, this
`neonatalbloodis a function ofgestational age and cannot
`would be adequate to measure functional SaOQg if MetHb
`be accurately predicted. HbF does produce a small error
`and COHbdid not absorb red or infrared light at the
`in in vitro laboratory oximeters; OgHbF may beinter-
`wavelengths used. Unfortunately,this is not the case, and
`preted as consisting partially of COHb.”°
`therefore both MetHb and COHb will cause errorsin the
`The absorbance ratio R (equation 4) may be affected
`pulse oximeter reading.It is not intuitively obvious how
`by any substancepresentin the pulsatile blood that absorbs
`a pulse oximeter will behave in the presence of dyshem-
`light at 660 or 940 nm andwasnotpresentin the same
`oglobins. With respect to carboxyhemoglobin, we can gain
`concentration in the volunteers used to generate the cal-
`some insight from the extinction curves of figure 2. In
`ibration curve(fig. 4). Intravenous dyes provide a good
`the infrared range (940 nm), COHbabsorbsverylittle
`exampleofthis principle.?!"*? Schelleret al. evaluated the
`light; whereas, in the red range (660 nm), it absorbs as
`effects of bolus doses of methylene blue, indigo carmine,
`muchlight as does OgHb. Thisis clinically illustrated by
`and indocyanine green on pulse oximeters in humanvol-
`the fact that patients with carboxyhemoglobinemia appear
`unteers.”’ They found that methylene blue causedafall
`red. Therefore, to the pulse oximeter, COHblookslike
`in SpO, to approximately 65% for 1-2 min. Indigo car-
`O2Hb at 660 nm; while, at 940 nm COHbisrelatively
`mine produceda very small dropin saturation, while in-
`transparent. Theeffect of COHbon pulse oximeter values
`docyanine green had an intermediate effect. The detec-
`has beenevaluated experimentally in dogs.’” In this study,
`tion of intravenous dyes by pulse oximeters should not
`the pulse oximeter saturation (SpO2) was foundto be given
`be surprising, because it was this effect that led Aoyagi
`approximately by:
`to the invention of pulse oximetry.”
`
`_ OgHb + 0.9 COHb
`SpO2 =
`total Hb
`
`x 100%.
`
`(8)
`
`Theeffects of methemoglobinemia on pulse oximetry
`are also partially predictable from the extinction curves
`(fig. 2). MetHb hasnearly the same absorbance as reduced
`hemoglobin at 660 nm,while it has a greater absorbance
`than the other hemoglobins at 940 nm. This is consistent
`with the clinical observation that methemoglobinemia
`produces very dark, brownish blood. Thus, it would be
`expected to producea large pulsatile absorbancesignal
`at both wavelengths. The effect of MetHb on pulse ox-
`imeter readings has also been measured in dogs.'* As
`methemoglobin levels increased, the pulse oximetersat-
`uration (SpOg) tended toward 85% and eventually became
`almost independentof the actual SaO,.'* In other words,
`in the presence of high levels of MetHb, SpOzis erro-
`neously low when SaQOzis above 85%, and erroneously
`high when SaQgis below 85%. This may be explained by
`the fact that MetHb causesa large pulsatile absorbance
`at both wavelengths, thereby adding to both the numer-
`ator and denominatorof the absorbanceratio R (equation
`4) and forcingthis ratio toward unity. As shownin figure
`
`LED CENTER WAVELENGTH VARIABILITY
`
`The LEDs used in pulse oximeter sensors are notideal
`monochromatic light sources; there is a narrow spectral
`range over which they emitlight. The center wavelength
`of the emission spectrum varies even amongdiodesofthe
`same type from the same manufacturer. This variation
`can be +15 nanometers.'® Asseen in figure 2, a shift in
`LED center wavelength will change the measured ex-
`tinction coefficient and thus produceanerrorin the sat-
`uration estimate. This source wavelength effect will be
`greatest for the red (660 nm) wavelength, because the
`extinction curves have a steeperslopeat this wavelength.
`Manufacturers have found two approaches to this prob-
`lem. Sometest all the LEDs andreject those that are out
`of their specified wavelength range, e.g., 660 + 5 nano-
`meters. This is expensive due to the number of LEDs
`
`§ Pologe JA, Raley DM: Effects of fetal hemoglobin on pulse ox-
`imetry. J Perinat VI1I:324-526, 1987.
`{ Anderson JV: The accuracy ofpulse oximetry in neonates: Effects
`of fetal hemoglobin andbilirubin. J Perinat VII:323, 1987.
`
`12
`
`12
`
`

`

`Anesthesiology
`V 70, No 1, Jan 1989
`
`PULSE OXIMETRY
`
`103
`
`rejected; i.e., narrower acceptable rangeyields improved
`accuracy but also more rejected LEDs. Alternatively,
`ot