Fetal pain a systematic multidisciplinary review

1. Fetal Pain: A Systematic Multidisciplinary Review of the
Evidence
Susan J. Lee; Henry J. Peter Ralston; Eleanor A. Drey; et al.
Online article and related content
current as of February 20, 2010. JAMA. 2005;294(8):947-954 (doi:10.1001/jama.294.8.947)
http://jama.ama-assn.org/cgi/content/full/294/8/947
Correction Contact me if this article is corrected.
Citations This article has been cited 28 times.
Contact me when this article is cited.
Topic collections Medical Practice; Health Policy; Law and Medicine; Anesthesia; Pain;
Patient-Physician Relationship/ Care; Patient-Physician Communication; Women's
Health; Pregnancy and Breast Feeding
Contact me when new articles are published in these topic areas.
Related Letters Fetal Pain
Laura B. Myers et al. JAMA. 2006;295(2):159.
Bobbi J. Lyman. JAMA. 2006;295(2):159.
Brian D. Sites. JAMA. 2006;295(2):160.
In Reply:
Susan J. Lee et al. JAMA. 2006;295(2):160.
Subscribe Email Alerts
http://jama.com/subscribe http://jamaarchives.com/alerts
Permissions Reprints/E-prints
permissions@ama-assn.org reprints@ama-assn.org
http://pubs.ama-assn.org/misc/permissions.dtl
Downloaded from www.jama.com by guest on February 20, 2010

2. CLINICAL REVIEW CLINICIAN’S CORNER
Fetal Pain
A Systematic Multidisciplinary Review of the Evidence
Susan J. Lee, JD Context Proposed federal legislation would require physicians to inform women
Henry J. Peter Ralston, MD seeking abortions at 20 or more weeks after fertilization that the fetus feels pain
and to offer anesthesia administered directly to the fetus. This article examines
Eleanor A. Drey, MD, EdM
whether a fetus feels pain and if so, whether safe and effective techniques exist for
John Colin Partridge, MD, MPH providing direct fetal anesthesia or analgesia in the context of therapeutic proce-
Mark A. Rosen, MD dures or abortion.
Evidence Acquisition Systematic search of PubMed for English-language articles
O
VER THE LAST SEVERAL
focusing on human studies related to fetal pain, anesthesia, and analgesia. Included
years, many states, includ- articles studied fetuses of less than 30 weeks’ gestational age or specifically addressed
ing California, Kentucky, fetal pain perception or nociception. Articles were reviewed for additional references.
Minnesota, Montana, New The search was performed without date limitations and was current as of June 6,
York, Oregon, and Virginia, have con- 2005.
sidered legislation requiring physi- Evidence Synthesis Pain perception requires conscious recognition or awareness
cians to inform women seeking abor- of a noxious stimulus. Neither withdrawal reflexes nor hormonal stress responses to
tions that the fetus feels pain and to offer invasive procedures prove the existence of fetal pain, because they can be elicited by
fetal anesthesia. This year, Arkansas and nonpainful stimuli and occur without conscious cortical processing. Fetal awareness
Georgia enacted such statutes.1,2 Cur- of noxious stimuli requires functional thalamocortical connections. Thalamocortical fi-
rently, Congress is considering legisla- bers begin appearing between 23 to 30 weeks’ gestational age, while electroencepha-
tion requiring physicians to inform lography suggests the capacity for functional pain perception in preterm neonates prob-
ably does not exist before 29 or 30 weeks. For fetal surgery, women may receive general
women seeking abortions 20 or more anesthesia and/or analgesics intended for placental transfer, and parenteral opioids
weeks after fertilization (ie, 22 weeks’ may be administered to the fetus under direct or sonographic visualization. In these
gestational age) that the fetus has “physi- circumstances, administration of anesthesia and analgesia serves purposes unrelated
cal structures necessary to experience to reduction of fetal pain, including inhibition of fetal movement, prevention of fetal
pain,” as evidenced by “draw[ing] away hormonal stress responses, and induction of uterine atony.
from surgical instruments.” The physi- Conclusions Evidence regarding the capacity for fetal pain is limited but indicates
cian must also offer anesthesia or anal- that fetal perception of pain is unlikely before the third trimester. Little or no evidence
gesia “administered directly” to the fe- addresses the effectiveness of direct fetal anesthetic or analgesic techniques. Simi-
tus. Physicians who do not comply may larly, limited or no data exist on the safety of such techniques for pregnant women in
be subject to substantial fines, license re- the context of abortion. Anesthetic techniques currently used during fetal surgery are
vocation, and civil suits for punitive not directly applicable to abortion procedures.
damages.3 JAMA. 2005;294:947-954 www.jama.com
Although this legislation would not
affect most US abortions because only
step in answering these questions, we Author Affiliations: School of Medicine (Ms Lee),
1.4% are performed at or after 21 weeks’ Department of Anatomy and W. M. Keck Founda-
reviewed the literature on fetal pain and
gestational age,4 this legislation raises tion for Integrative Neuroscience (Dr Ralston), and
fetal anesthesia and analgesia. Departments of Obstetrics, Gynecology and Repro-
important scientific, clinical, ethical, ductive Sciences (Drs Drey and Rosen), Pediatrics
and policy issues. When does a fetus EVIDENCE ACQUISITION (Dr Partridge), and Anesthesia and Perioperative
Care (Dr Rosen), University of California, San Fran-
have the functional capacity to feel English-language articles involving hu- cisco.
pain? If that capacity exists, what forms man participants were searched using Corresponding Author: Mark A. Rosen, MD, Depart-
ment of Anesthesia and Perioperative Care, Univer-
of anesthesia or analgesia are safe and PubMed for (1) fetal pain (16 articles), sity of California, San Francisco, 513 Parnassus Ave,
effective for treating fetal pain? As a first fetal anesthesia (6 articles), and fetal an- San Francisco, CA 94143-0648 (rosenm@anesthesia
.ucsf.edu).
algesia (3 articles); (2) fetus and (anes- Clinical Review Section Editor: Michael S. Lauer, MD.
CME available online at thesia or analgesia) (1239 articles); (3) We encourage authors to submit papers for consider-
www.jama.com ation as a “Clinical Review.” Please contact Michael S.
Medical Subject Headings (MeSH) an- Lauer, MD, at lauerm@ccf.org.
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, August 24/31, 2005—Vol 294, No. 8 947
Downloaded from www.jama.com by guest on February 20, 2010

3. FETAL PAIN
algesics/administration and dosage and fe-
Figure. Spinal Reflex and Pain Perception Pathways
tus (44 articles); (4) MeSH anesthesia/
administration and dosage and fetus (0
A Spinal Reflex articles); (5) (neurodevelopment or devel-
opment or anatomy) and (fetus or fetal)
Dorsal Root Ganglion SPINAL CORD
and (pain or nociception or noxious) (306
articles); (6) (thalamocortical or thala-
2 Peripheral 3 Dorsal Horn mus or cortex) and (fetus or fetal) and
Sensory Neuron Interneuron
(pain or nociception or noxious) (13 ar-
ticles); (7) (electroencephalog* or EEG or
1 Noxious evoked potential) and (fetus or fetal or pre-
Stimulus mature neonate or premature infant or pre-
4 Ventral Horn term neonate or preterm infant) and (pain
L E
Motor Neuron or nociception or noxious or conscious*)
SC
MU (7 articles); (8) fetal and pain and (re-
5 Contraction sponse or assessment or facial expres-
sion) (112 articles); and (9) facial expres-
sion and (fetus or fetal) or ([neonate or
neonatal or infant] and [premature or pre-
B Pain Perception via the Spinothalamic Tract term]) and (pain or nociception or nox-
ious) (360 articles). The search was per-
C
BRAIN formed without date limitations and was
O
R
current as of June 6, 2005. From these
TE
X
6 Thalamocortical Axon
7 Perception
search results, we excluded articles that
did not study fetuses of less than 30
weeks’ gestational age or that did not spe-
5 Thalamus cifically address fetal pain perception or
nociception. With a focus on topics ad-
dressed by earlier review articles on fe-
tal pain, anesthesia, and analgesia, ar-
ticles were reviewed for additional
references.
EVIDENCE SYNTHESIS
4 Spinothalamic What Is Pain?
3 Spinothalamic Tract
Pain is a subjective sensory and emo-
Neuron
tional experience that requires the pres-
Dorsal Root Ganglion ence of consciousness to permit recog-
nition of a stimulus as unpleasant.5-7
2 Peripheral
Sensory Neuron
Although pain is commonly associated
with physical noxious stimuli, such as
when one suffers a wound, pain is fun-
damentally a psychological construct that
1 Noxious
Stimulus
may exist even in the absence of physi-
cal stimuli, as seen in phantom limb
SPINAL CORD
pain.5,7 The psychological nature of pain
also distinguishes it from nociception,
which involves physical activation of no-
ciceptive pathways without the subjec-
A, Reflex responses to noxious stimuli occur early in development, before thalamocortical circuits are func-
tional; noxious stimuli trigger reflex movement without cortical involvement. Activated by a noxious stimulus
tive emotional experience of pain.5,8 For
(1), a peripheral sensory neuron (2) synapses on a dorsal horn interneuron (3) that in turn synapses on a ven- example, nociception without pain ex-
tral horn motor neuron (4), leading to reflex muscle contraction and limb withdrawal (5). B, Later in develop- ists below the level of a spinal cord le-
ment, noxious stimuli (1) activate peripheral sensory neurons (2) that synapse on spinothalamic tract neurons
(3), the axons of which extend up the spinal cord as the spinothalamic tract (4) to synapse on neurons of the sion, where reflex withdrawal from a
thalamus (5). From here, thalamocortical axons synapse on cortical neurons, resulting in the conscious per- noxious stimulus occurs without con-
ception of pain.
scious perception of pain (FIGURE, A).5
948 JAMA, August 24/31, 2005—Vol 294, No. 8 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on February 20, 2010

4. FETAL PAIN
Table. Anatomical and Functional Development of Nociception and Pain Perception Pathways
Anatomical/ Gestational
Functional Characteristic Description Age, wk Source
Peripheral cutaneous sensory receptors Perioral cutaneous sensory receptors 7.5
Palmar cutaneous sensory receptors 10-10.5 Humphrey,13 1964
Abdominal cutaneous sensory receptors 15
Spinal cord Spinal reflex arc in response to nonnoxious stimuli 8 Okado and Kojima,14 1984
Neurons for nociception in dorsal root ganglion 19 Konstantinidou et al,15 1995
Thalamic afferents Thalamic afferents reach subplate zone 20-22 Kostovic and Rakic,16 1990
Hevner,17 2000
Thalamic afferents reach cortical plate 23-24 Kostovic and Rakic,18 1984
Kostovic and Goldman-Rakic,19 1983
Cortical function* Somatosensory evoked potentials with distinct, 29 Klimach and Cooke,20 1988
constant components Hrbek et al,21 1973
First electrocardiographic pattern denoting both 30 Clancy et al,22 2003
wakefulness and active sleep Torres and Anderson,23 1985
*Earliest evidence of functional thalamocortical connections required for conscious perception of pain.
Because pain is a psychological con- cortical pathways required for con- showing initial cortical plate penetra-
struct with emotional content, the ex- scious perception of pain (TABLE). tion at 22 weeks’ developmental age (24
perience of pain is modulated by chang- No human studies have directly ex- weeks’ gestational age).24
ing emotional input and may need to be amined the development of thalamo- In a study of 8 human fetuses, me-
learned through life experience.7,9,10 Re- cortical circuits associated with pain diodorsal thalamic afferents were first
gardless of whether the emotional con- perception. The developmental age at observed in the cortical plate at 22
tent of pain is acquired, the psychologi- which thalamic pain fibers reach the weeks’ developmental age (24 weeks’
cal nature of pain presupposes the cortex has been inferred from studies gestational age).19 While connections
presence of functional thalamocortical of other thalamocortical circuits, which between mediodorsal afferents and the
circuitry required for conscious percep- may or may not develop at the same anterior cingulate cortex25 may be rel-
tion, as discussed below. time as thalamic fibers mediating cor- evant to pain perception,12,26 this study
tical perception of pain. examined mediodorsal afferents to un-
Fetal Capacity for Pain These histological neurodevelop- specified regions of the frontal cor-
Neuroanatomy and Development. No- ment studies typically describe fetal ma- tex,19 which serves numerous func-
ciception may be characterized by re- turity in terms of developmental age, tions unrelated to pain perception.19,27
flex movement in response to a nox- representing the number of weeks post- Another histological study of 12 speci-
ious stimulus, without cortical ovulation or postfertilization. Clini- mens found that afferents from unspeci-
involvement or conscious pain percep- cians regularly use gestational age, rep- fied thalamic regions reached the devel-
tion. Nociception involves peripheral resenting weeks from the first day of the oping prefrontal cortex in 1 preterm
sensory receptors whose afferent fi- woman’s last menstrual period. When neonate of 27 weeks’ developmental age,
bers synapse in the spinal cord on in- referring to a fetus at the same point in concluding that thalamic fibers begin
terneurons, which synapse on motor development, the gestational age is ap- entering the cortex between 26 and 28
neurons that also reside in the spinal proximately 2 weeks greater than the weeks’ developmental age (28 and 30
cord. These motor neurons trigger developmental age. weeks’ gestational age).28 A different study
muscle contraction, causing limb flex- A histological study of the visual found that thalamic afferents had not
ion away from a stimulus (Figure, A).11 pathway in 8 human fetuses, each at a reached the somatosensory cortical plate
In contrast, pain perception re- different developmental age, con- by 22 weeks’ developmental age (24
quires cortical recognition of the stimu- cluded that thalamic projections reach weeks’ gestational age). By 24 weeks’
lus as unpleasant. Peripheral sensory re- the visual cortex at 21 to 25 weeks’ de- developmental age (26 weeks’ gesta-
ceptor afferents synapse on spinal cord velopmental age (approximately 23-27 tional age), the density of cortical plate
neurons, the axons of which project to weeks’ gestational age), based on re- synapses increased, although these were
the thalamus, which sends afferents to sults from a fetus of 24 weeks’ devel- not necessarily from thalamic affer-
the cerebral cortex (Figure, B),11 acti- opmental age (26 weeks’ gestational ents.16 Based on these studies, direct thala-
vating any number of cortical re- age).18 A similar 7-fetus study found mocortical fibers that are not specific for
gions.12 Sensory receptors and spinal thalamic afferents reached the audi- pain begin to emerge between 21 and 28
cord synapses required for nocicep- tory cortical plate at 24 to 26 weeks’ de- weeks’ developmental age (23 and 30
tion develop earlier than the thalamo- velopmental age, with 1 specimen weeks’ gestational age).
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, August 24/31, 2005—Vol 294, No. 8 949
Downloaded from www.jama.com by guest on February 20, 2010

5. FETAL PAIN
However, others have proposed that 24 weeks’ postconceptional age (PCA) PCA, suggesting that the capacity of the
thalamocortical connections could also (ie, the gestational age plus number of neonate to distinguish between nox-
be established indirectly if thalamic af- weeks postpartum).22 Electroencepha- ious and nonnoxious stimuli is matur-
ferents were to synapse on subplate neu- lographic activity is normally asynchro- ing. 37 Furthermore, flexion with-
rons, which could synapse on cortical nous between the hemispheres and drawal from tactile stimuli is a
plate neurons.29 The subplate is a tran- mostly discontinuous at less than 27 noncortical spinal reflex exhibited by
sient fetal structure 1 layer deep to the weeks’ PCA,23,33,34 becoming mostly infants with anencephaly38 and by in-
cortical plate and serves as a “waiting continuous around 34 weeks’ PCA.23,34 dividuals in a persistent vegetative
compartment” for various afferents, in- Interhemispheric synchrony increases state39 who lack cortical function.
cluding thalamic afferents, en route to around 29 to 30 weeks’ PCA, then de- Behavioral studies have also identi-
the cortical plate.16,29,30 The subplate re- clines, then increases again, reaching al- fied a distinct set of neonatal facial
cedes after 30 weeks’ developmental most complete synchrony by term.22,33 movements present during invasive
age,16,29 while the cortical plate ma- Given these baseline differences be- procedures such as heel lancing but
tures into the 6 layers of the cerebral tween neonatal and adult EEGs, pat- absent during noninvasive proce-
cortex.28 In contrast to direct thalamo- terns associated with impaired con- dures.40-46 These facial movements,
cortical fibers, which are not visible un- sciousness in adults33,35 are inapplicable which are similar to those of adults ex-
til almost the third trimester, thalamic to the analysis of neonatal EEGs. periencing pain,47,48 were evident in neo-
afferents begin to reach the somatosen- Some investigators contend that EEG nates at 28 to 30 weeks’ PCA but not
sory subplate at 18 weeks’ developmen- patterns denoting wakefulness indi- at 25 to 27 weeks’ PCA.40 Facial move-
tal age (20 weeks’ gestational age)16 and cate when consciousness is first pos- ments may not necessarily be corti-
the visual subplate at 20 to 22 weeks’ sible.5,36 Wakefulness is a state of arousal cally controlled.49 One study found no
gestational age.17 These afferents ap- mediated by the brainstem and thala- difference in facial activity during heel
pear morphologically mature enough to mus in communication with the lancing of neonates with and without
synapse with subplate neurons,31 al- cortex.5,22 In preterm neonates, the ear- significant cortical injury, suggesting
though no human study has shown that liest EEG pattern representing wake- that facial activity even around 32
functional synapses exist between tha- fulness appears around 30 weeks’ weeks’ PCA may not represent con-
lamic afferents and subplate neurons. PCA.22,23 However, wakefulness alone scious perception of pain.50
Subplate neurons may synapse with cor- is insufficient to establish conscious- Stress Responses. Hemodynamic
tical plate neurons and direct the growth ness, as unconscious patients in a per- and neuroendocrine changes in fe-
of thalamic afferents to their final syn- sistent vegetative state may also have tuses undergoing stressful procedures
aptic targets in the cortical plate.29 De- wakeful EEGs.5,36 have also been used to infer pain per-
spite this developmental role, no hu- Somatosensory evoked potentials ception.51 As early as 16 weeks’ gesta-
man study has shown that synapses (SEPs) may also provide evidence of tional age, fetal cerebral blood flow in-
between subplate and cortical plate neu- pain processing in the somatosensory creases during venipuncture and
rons convey information about pain per- cortex, although they are not used clini- transfusions that access the fetal he-
ception from the thalamus to the de- cally to test pain pathways. SEPs test the patic vein through the innervated fe-
veloping cortex. dorsal column tract of the spinal cord, tal abdominal wall but not during ve-
Electroencephalography. The his- which transmits visceral pain sensa- nipuncture and transfusions involving
tological presence of thalamocortical tion to the somatosensory cortex via the the noninnervated umbilical cord.52 In-
fibers is insufficient to establish capac- thalamus.12 SEPs with distinct and con- creased cerebral blood flow is not nec-
ity for pain perception. These anatomi- stant N1 components of normal peak essarily indicative of pain, as this re-
cal structures must also be functional. latency are present at 29 weeks’ PCA, sponse is thought to constitute a “brain
Although no electroencephalographic indicating that thalamic connections sparing” mechanism associated with
“pain pattern” exists, electroencepha- with the somatosensory cortex are func- hypoxia 53 and intrauterine growth
lography may be one way of assessing tional at that time.20,21 restriction.54
general cortical function because elec- Behavioral Studies. Although widely Other investigators measured in-
troencephalograms (EEGs) measure used to assess pain in neonates, with- creases in fetal plasma concentrations of
summated synaptic potentials from cor- drawal reflexes and facial movements cortisol, -endorphin, and noradrena-
tical neurons. However, EEG activity do not necessarily represent con- line associated with intrauterine nee-
alone does not prove functionality, be- scious perception of pain. Full-term dling procedures, finding that increases
cause neonates with anencephaly who neonates exhibit a “cutaneous with- during blood sampling from the hepatic
lack functional neural tissue above the drawal reflex” that is activated at a vein were greater than those during sam-
brainstem may still have EEG activity.32 threshold much lower than that which pling from the umbilical cord.55,56 How-
Normal EEG patterns have been would produce discomfort in a child or ever, these neuroendocrine responses do
characterized for neonates as young as adult.37 This threshold increases with not constitute evidence of fetal pain,
950 JAMA, August 24/31, 2005—Vol 294, No. 8 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on February 20, 2010

6. FETAL PAIN
because the autonomic nervous system ward the fetus represents the chief jus- general anesthesia are unjustified for
and hypothalamic-pituitary-adrenal axis tification for using fetal anesthesia or these procedures; adults typically un-
mediate them without conscious corti- analgesia during abortion—to relieve dergo similar procedures with no an-
cal processing.57 Additionally, these suffering if fetal pain exists. As with any algesia or only local analgesia.67 No es-
responses are not specific for painful clinical decision, thorough safety and tablished fetal analgesia protocol exists
stimuli. Plasma noradrenaline concen- risk-benefit analyses should be under- for these procedures, although 3 tech-
trations may increase after umbilical cord taken before performing an interven- niques have been proposed, namely, di-
transfusion,56 and plasma -endorphin tion. Because the principle of benefi- rect delivery of medications to the fe-
concentrations may increase after cence also requires the woman’s tus, delivery of medications to the fetus
repeated cordocenteses.58 Plasma corti- physician to act in her best interests, po- via maternal intravenous infusion, and
sol and -endorphin concentrations tential fetal benefit must be weighed intra-amniotic delivery of medications.
increase during innocuous activities such against real risks to the woman’s health. Direct Delivery. One group has ex-
as exercise.59 Moreover, in adults, neu- The safety and effectiveness of pro- amined the effects of analgesics deliv-
roendocrine stress responses may per- posed fetal anesthesia and analgesia ered directly to human fetuses during
sist despite well-controlled postopera- techniques are discussed below. minimally invasive procedures. 8 7
tive pain.60 General Anesthesia for Fetal Surgery. Twenty-eight fetuses that received in-
Vital signs also have been used to as- Fetal surgery involving laparotomy, hys- travenous fentanyl before hepatic vein
sess neonatal pain.42,43,45,51,61 However, terotomy, or both requires general or blood transfusions had diminished
heart rate, respiratory rate, and trans- regional anesthesia.68,76 Regional anes- changes in plasma -endorphin con-
cutaneous oxygen and carbon dioxide thesia, such as epidural anesthesia, does centration and cerebral blood flow, com-
levels do not necessarily differ signifi- not anesthetize the fetus.76 General anes- pared with fetuses not receiving fen-
cantly between alcohol-swabbing and thesia is more commonly used because tanyl. The cortisol response was not
lancing the heels of preterm neo- it induces uterine atony and fetal immo- significantly decreased with fentanyl.
nates.40 Another group found that a simi- bilization.65,77 Studies of inhalational The investigators did not examine risks
lar proportion of neonates became hy- agents in pregnant ewes determined that for the woman, such as infection or un-
poxic during tracheal suction, as well as a dose capable of anesthetizing the ewe controlled bleeding.76 Furthermore, re-
during nonnoxious routine care such as also anesthetized the fetus.78 Admin- ducing the stress response is distinct
washing and weighing.62 istering fentanyl, pancuronium, or from reducing pain. For example,
vecuronium to the fetus intramuscu- plasma glucose and cortisol concentra-
Fetal Anesthesia and Analgesia larly may supplement analgesia or tions may not differ significantly be-
Anesthetics and analgesics are com- immobilization.64,65,77,79 tween adults with and without postop-
monly used to alleviate pain and dis- For pregnant women, general anes- erative pain.60
comfort. Despite ongoing debate re- thesia is associated with increased mor- Delivery via Maternal Intravenous
garding fetal capacity for pain, fetal bidity and mortality, particularly because Infusion. To achieve presumably effec-
anesthesia and analgesia are still war- of airway-related complications80-82 and tive fetal plasma concentrations of fen-
ranted for surgical procedures under- increased risk of hemorrhage from uter- tanyl by placental transfer, potentially
taken to promote fetal health. When ine atony.70 Historically, general anes- unsafe doses would need to be admin-
long-term fetal well-being is a central thesia was used in abortions, even in the istered to the woman.88 Although stan-
consideration, evidence of fetal pain is first trimester, until studies found that dard doses of fentanyl are generally safe
unnecessary to justify fetal anesthesia general anesthesia was a leading cause for maternal analgesia during labor,89
and analgesia because they serve other of abortion-related mortality.83-85 In addi- fentanyl can pose serious risks such as
purposes unrelated to pain reduction, tion to safety concerns, general anes- hypoventilation if maternal doses are
including (1) inhibiting fetal move- thesia increases the cost of abortion, significantly increased to achieve more
ment during a procedure 63-65 ; (2) making it prohibitively expensive for the extensive placental transfer.67,68 Se-
achieving uterine atony to improve sur- majority of patients who pay out of vere maternal hypoventilation may re-
gical access to the fetus and to prevent pocket.86 quire endotracheal intubation, which
contractions and placental separa- Anesthesia and Analgesia in Mini- increases risks and costs for the woman,
tion66-70; (3) preventing hormonal stress mally Invasive Fetal Procedures. In as described above.
responses associated with poor surgi- contrast to fetal surgery requiring re- No data exist on the dosing or effi-
cal outcomes in neonates71,72; and (4) gional or general anesthesia, mini- cacy of using medications such as di-
preventing possible adverse effects on mally invasive fetal procedures do not azepam and morphine for fetal analge-
long-term neurodevelopment and be- involve maternal laparotomy or hys- sia via maternal intravenous infusion,
havioral responses to pain.73-75 terotomy and instead use needles or en- although studies have characterized the
These objectives are not applicable doscopy to access the fetus. For the sake placental transfer of these medica-
to abortions. Instead, beneficence to- of reducing pain, the increased risks of tions.90-92 Two related studies found that
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, August 24/31, 2005—Vol 294, No. 8 951
Downloaded from www.jama.com by guest on February 20, 2010

7. FETAL PAIN
low-dose remifentanil via maternal in- these tests of cortical function suggest Because pain perception probably
travenous infusion achieved fetal im- that conscious perception of pain does does not function before the third tri-
mobilization during laser coagulation not begin before the third trimester. mester, discussions of fetal pain for
of placental vessels.93,94 However, im- Cutaneous withdrawal reflexes and abortions performed before the end of
mobilization is not the equivalent of hormonal stress responses present ear- the second trimester should be non-
pain reduction, and these procedures lier in development are not explicit or compulsory. Fetal anesthesia or anal-
did not involve surgery on the fetus. sufficient evidence of pain perception gesia should not be recommended or
Intra-amniotic Delivery. Intra- because they are not specific to nox- routinely offered for abortion because
amniotic injection would be techni- ious stimuli and are not cortically current experimental techniques pro-
cally simpler than direct fetal injec- mediated. vide unknown fetal benefit and may in-
tion, although the drug must be A variety of anesthetic and analge- crease risks for the woman. Instead, fur-
absorbed through fetal membranes and sic techniques have been used for fetal ther research should focus on when
skin. Intra-amniotic sufentanil injec- surgery, including maternal general an- pain-related thalamocortical path-
tion in 10 pregnant ewes resulted in fe- esthesia, regional anesthesia, and ad- ways become functional in humans. If
tal plasma concentrations that would ministration of medications for placen- the fetus can feel pain, additional re-
control postoperative pain in human tal transfer to the fetus. However, these search may lead to effective fetal anes-
adults.95,96 Sufentanil concentrations in techniques are not necessarily appli- thesia or analgesia techniques that are
the ewes also reached adult human cable to abortions. Surgical proce- also safe for women.
therapeutic concentrations without dures undertaken for fetal benefit use Financial Disclosures: None reported.
causing significant hemodynamic anesthesia to achieve objectives unre-
changes.96 However, the study did not lated to pain control, such as uterine REFERENCES
evaluate fetal response to noxious relaxation, fetal immobilization, and 1. Unborn Child Pain Awareness and Prevention Act
stimuli, and no data exist regarding possible prevention of neuroendo- of 2005. To be codified at Ark Code Ann §§20-16-
1101 to 1111.
safety or effectiveness in humans. crine stress responses associated with 2. Woman’s Right to Know Act. To be codified at Ga
poor surgical outcomes. Thus, fetal an- Code Ann §31-9A-4.
CONCLUSIONS esthesia may be medically indicated for 3. Unborn Child Pain Awareness Act, S51, 109th Cong
(2005).
Pain is an emotional and psychologi- fetal surgery regardless of whether fe- 4. Strauss LT, Herndon J, Chang J, et al. Abortion sur-
cal experience that requires conscious tal pain exists. veillance—United States, 2001. MMWR Surveill
Summ. 2004;53:1-32.
recognition of a noxious stimulus. Con- In the context of abortion, fetal an- 5. Benatar D, Benatar M. A pain in the fetus: toward
sequently, the capacity for conscious algesia would be used solely for benefi- ending confusion about fetal pain. Bioethics. 2001;15:
57-76.
perception of pain can arise only after cence toward the fetus, assuming fetal 6. Glover V, Fisk NM. Fetal pain: implications for re-
thalamocortical pathways begin to func- pain exists. This interest must be con- search and practice. Br J Obstet Gynaecol. 1999;106:
tion, which may occur in the third tri- sidered in concert with maternal safety 881-886.
7. International Association for the Study of Pain.
mester around 29 to 30 weeks’ gesta- and fetal effectiveness of any pro- IASP Pain Terminology. 2004. Available at: http:
tional age, based on the limited data posed anesthetic or analgesic tech- //www.iasp-pain.org/terms-p.html. Accessed May 2,
2005.
available. Small-scale histological stud- nique. For instance, general anesthe- 8. Anand KJ, Hickey PR. Pain and its effects in the hu-
ies of human fetuses have found that sia increases abortion morbidity and man neonate and fetus. N Engl J Med. 1987;317:1321-
1329.
thalamocortical fibers begin to form be- mortality for women and substan- 9. Derbyshire SW. Locating the beginnings of pain.
tween 23 and 30 weeks’ gestational age, tially increases the cost of abortion. Al- Bioethics. 1999;13:1-31.
but these studies did not specifically ex- though placental transfer of many opi- 10. Derbyshire SW. Fetal pain: an infantile debate.
Bioethics. 2001;15:77-84.
amine thalamocortical pathways ac- oids and sedative-hypnotics has been 11. Fitzgerald M, Howard RF. The neurobiologic
tive in pain perception. determined, the maternal dose re- basis of pediatric pain. In: Schechter NL, Berde CB,
Yaster M, eds. Pain in Infants, Children, and Adoles-
While the presence of thalamocorti- quired for fetal analgesia is unknown, cents. 2nd ed. Philadelphia, Pa: Lippincott Williams &
cal fibers is necessary for pain per- as is the safety for women at such doses. Wilkins; 2003:19-42.
12. Strigo IA, Duncan GH, Boivin M, Bushnell MC.
ception, their mere presence is in- Furthermore, no established proto- Differentiation of visceral and cutaneous pain in the
sufficient—this pathway must also be cols exist for administering anesthesia human brain. J Neurophysiol. 2003;89:3294-3303.
13. Humphrey T. Some correlations between the
functional. It has been proposed that or analgesia directly to the fetus for appearance of fetal reflexes and the development
transient, functional thalamocortical minimally invasive fetal procedures or of the nervous system. Prog Brain Res. 1964;4:93-
circuits may form via subplate neu- abortions. Experimental techniques, 135.
14. Okado N, Kojima T. Ontogeny of the central ner-
rons around midgestation, but no such as administration of fentanyl di- vous system: neurogenesis, fibre connection, synap-
human study has demonstrated this rectly to the fetus and intra-amniotic togenesis and myelination in the spinal cord. In: Prechtl
HFR, ed. Clinics in Developmental Medicine: Conti-
early functionality. Instead, constant injection of sufentanil in pregnant nuity of Neural Functions From Prenatal to Postna-
SEPs appear at 29 weeks’ PCA, and ewes, have not been shown to de- tal Life. Vol 94. Philadelphia, Pa: JB Lippincott Co; 1984:
31-45.
EEG patterns denoting wakefulness crease fetal pain and are of unknown 15. Konstantinidou AD, Silos-Santiago I, Flaris N, Snider
appear around 30 weeks’ PCA. Both of safety in humans. WD. Development of the primary afferent projection
952 JAMA, August 24/31, 2005—Vol 294, No. 8 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on February 20, 2010

8. FETAL PAIN
in human spinal cord. J Comp Neurol. 1995;354: patterns. In: Niedermeyer E, Lopes da Silva FH, eds. 56. Giannakoulopoulos X, Teixeira J, Fisk N, Glover
11-12. Electroencephalography: Basic Principles, Clinical Ap- V. Human fetal and maternal noradrenaline re-
16. Kostovic I, Rakic P. Developmental history of the plications, and Related Fields. 4th ed. Philadelphia, sponses to invasive procedures. Pediatr Res. 1999;45:
transient subplate zone in the visual and somatosen- Pa: Lippincott Williams & Wilkins; 1999:215-234. 494-499.
sory cortex of the macaque monkey and human brain. 36. Burgess JA, Tawia SA. When did you first begin 57. Carrasco GA, Van de Kar LD. Neuroendocrine
J Comp Neurol. 1990;297:441-470. to feel it?—locating the beginning of human pharmacology of stress. Eur J Pharmacol. 2003;463:
17. Hevner RF. Development of connections in the consciousness. Bioethics. 1996;10:1-26. 235-272.
human visual system during fetal mid-gestation: a DiI- 37. Andrews K, Fitzgerald M. The cutaneous with- 58. Radunovic N, Lockwood CJ, Ghidini A, Alvarez
tracing study. J Neuropathol Exp Neurol. 2000;59:385- drawal reflex in human neonates: sensitization, re- M, Berkowitz RL. Is fetal blood sampling associated
392. ceptive fields, and the effects of contralateral with increased beta-endorphin release into the fetal
18. Kostovic I, Rakic P. Development of prestriate vi- stimulation. Pain. 1994;56:95-101. circulation? Am J Perinatol. 1993;10:112-114.
sual projections in the monkey and human fetal ce- 38. Ashwal S, Peabody JL, Schneider S, Tomasi LG, 59. Williams RH, Larsen PR. Williams Textbook of
rebrum revealed by transient cholinesterase staining. Emery JR, Peckham N. Anencephaly: clinical determi- Endocrinology. 10th ed. Philadelphia, Pa: Saunders;
J Neurosci. 1984;4:25-42. nation of brain death and neuropathologic studies. Pe- 2003.
19. Kostovic I, Goldman-Rakic PS. Transient cholin- diatr Neurol. 1990;6:233-239. 60. Schulze S, Roikjaer O, Hasselstrom L, Jensen NH,
esterase staining in the mediodorsal nucleus of the 39. Pilon M, Sullivan SJ. Motor profile of patients in Kehlet H. Epidural bupivacaine and morphine plus sys-
thalamus and its connections in the developing hu- minimally responsive and persistent vegetative states. temic indomethacin eliminates pain but not systemic
man and monkey brain. J Comp Neurol. 1983;219: Brain Inj. 1996;10:421-437. response and convalescence after cholecystectomy.
431-447. 40. Craig KD, Whitfield MF, Grunau RV, Linton J, Had- Surgery. 1988;103:321-327.
20. Klimach VJ, Cooke RW. Maturation of the neo- jistavropoulos HD. Pain in the preterm neonate: be- 61. Porter FL, Wolf CM, Miller JP. Procedural pain in
natal somatosensory evoked response in preterm havioural and physiological indices. Pain. 1993;52:287- newborn infants: the influence of intensity and
infants. Dev Med Child Neurol. 1988;30:208-214. 299. development. Pediatrics. 1999;104:e13.
21. Hrbek A, Karlberg P, Olsson T. Development of 41. Johnston CC, Stevens BJ, Yang F, Horton L. Dif- 62. Pokela ML. Pain relief can reduce hypoxemia in
visual and somatosensory evoked responses in pre- ferential response to pain by very premature neonates. distressed neonates during routine treatment
term newborn infants. Electroencephalogr Clin Pain. 1995;61:471-479. procedures. Pediatrics. 1994;93:379-383.
Neurophysiol. 1973;34:225-232. 42. Johnston CC, Stevens B, Yang F, Horton L. De- 63. Seeds JW, Corke BC, Spielman FJ. Prevention of
22. Clancy RR, Bergqvist AGC, Dlugos DJ. Neonatal velopmental changes in response to heelstick in pre- fetal movement during invasive procedures with pan-
electroencephalography. In: Ebersole JS, Pedley TA, term infants: a prospective cohort study. Dev Med curonium bromide. Am J Obstet Gynecol. 1986;155:
eds. Current Practice of Clinical Electroencephalog- Child Neurol. 1996;38:438-445. 818-819.
raphy. 3rd ed. Philadelphia, Pa: Lippincott Williams & 43. Lindh V, Wiklund U, Sandman PO, Hakansson S. 64. Rosen MA. Anesthesia for fetal procedures and
Wilkins; 2003:160-234. Assessment of acute pain in preterm infants by evalu- surgery. Yonsei Med J. 2001;42:669-680.
23. Torres F, Anderson C. The normal EEG of the hu- ation of facial expression and frequency domain analy- 65. Cauldwell CB. Anesthesia for fetal surgery. An-
man newborn. J Clin Neurophysiol. 1985;2:89-103. sis of heart rate variability. Early Hum Dev. 1997;48: esthesiol Clin North America. 2002;20:211-226.
24. Krmpotic-Nemanic J, Kostovic I, Kelovic Z, Nemanic 131-142. 66. Rosen MA. Anesthesia for procedures involving
D, Mrzljak L. Development of the human fetal audi- 44. Hadjistavropoulos HD, Craig KD, Grunau RE, Whit- the fetus. Semin Perinatol. 1991;15:410-417.
tory cortex: growth of afferent fibres. Acta Anat (Basel). field MF. Judging pain in infants: behavioural, con- 67. Smith RP, Gitau R, Glover V, Fisk NM. Pain and
1983;116:69-73. textual, and developmental determinants. Pain. 1997; stress in the human fetus. Eur J Obstet Gynecol Re-
25. Vogt BA, Rosene DL, Pandya DN. Thalamic and 73:319-324. prod Biol. 2000;92:161-165.
cortical afferents differentiate anterior from posterior 45. Goubet N, Clifton RK, Shah B. Learning about pain 68. White MC, Wolf AR. Pain and stress in the hu-
cingulate cortex in the monkey. Science. 1979;204: in preterm newborns. J Dev Behav Pediatr. 2001;22: man fetus. Best Pract Res Clin Anaesthesiol. 2004;18:
205-207. 418-424. 205-220.
26. Schnitzler A, Ploner M. Neurophysiology and func- 46. Xia C, Yang L, Zhao P, Zhang X. Response to pain 69. Rosen MA. Anesthesia and tocolysis for fetal
tional neuroanatomy of pain perception. J Clin by different gestational age neonates. J Huazhong Univ intervention. In: Harrison MR, Golbus MS, Filly RA, eds.
Neurophysiol. 2000;17:592-603. Sci Technolog Med Sci. 2002;22:84-86. The Unborn Patient: Prenatal Diagnosis and Treat-
27. Barbas H. Connections underlying the synthesis 47. Craig KD, Prkachin KM, Grunau RE. Facial ex- ment. Orlando, Fla: Grune & Stratton; 1984:417-433.
of cognition, memory, and emotion in primate pression of pain. In: Turk DC, Melzack R, eds. Hand- 70. Schwarz U, Galinkin JL. Anesthesia for fetal surgery.
prefrontal cortices. Brain Res Bull. 2000;52:319- book of Pain Assessment. 2nd ed. New York, NY: Guil- Semin Pediatr Surg. 2003;12:196-201.
330. ford Press; 2001:153-169. 71. Anand KJ, Hickey PR. Halothane-morphine com-
28. Mrzljak L, Uylings HB, Kostovic I, Van Eden CG. 48. Craig KD, Hadjistavropoulos HD, Grunau RV, pared with high-dose sufentanil for anesthesia and
Prenatal development of neurons in the human pre- Whitfield MF. A comparison of two measures of fa- postoperative analgesia in neonatal cardiac surgery.
frontal cortex, I: a qualitative Golgi study. J Comp cial activity during pain in the newborn child. J Pedi- N Engl J Med. 1992;326:1-9.
Neurol. 1988;271:355-386. atr Psychol. 1994;19:305-318. 72. Anand KJ, Sippell WG, Aynsley-Green A. Ran-
29. Kostovic I, Judas M. Correlation between the se- 49. Waxman SG. Clinical observations on the emo- domised trial of fentanyl anaesthesia in preterm ba-
quential ingrowth of afferents and transient patterns tional motor system. Prog Brain Res. 1996;107:595-604. bies undergoing surgery: effects on the stress response.
of cortical lamination in preterm infants. Anat Rec. 50. Oberlander TF, Grunau RE, Fitzgerald C, Whit- Lancet. 1987;1:62-66.
2002;267:1-6. field MF. Does parenchymal brain injury affect biobe- 73. Johnston CC, Stevens BJ. Experience in a neona-
30. Ulfig N, Neudorfer F, Bohl J. Transient structures havioral pain responses in very low birth weight in- tal intensive care unit affects pain response. Pediatrics.
of the human fetal brain: subplate, thalamic reticular fants at 32 weeks’ postconceptional age? Pediatrics. 1996;98:925-930.
complex, ganglionic eminence. Histol Histopathol. 2002;110:570-576. 74. Taddio A, Katz J, Ilersich AL, Koren G. Effect of
2000;15:771-790. 51. Franck LS, Miaskowski C. Measurement of neo- neonatal circumcision on pain response during sub-
31. Kostovic I, Judas M, Petanjek Z, Simic G. Onto- natal responses to painful stimuli: a research review. sequent routine vaccination. Lancet. 1997;349:
genesis of goal-directed behavior: anatomo- J Pain Symptom Manage. 1997;14:343-378. 599-603.
functional considerations. Int J Psychophysiol. 1995; 52. Teixeira JM, Glover V, Fisk NM. Acute cerebral 75. Taylor A, Fisk NM, Glover V. Mode of delivery
19:85-102. redistribution in response to invasive procedures in the and subsequent stress response. Lancet. 2000;
32. Schenk VW, De Vlieger M, Hamersma K, De human fetus. Am J Obstet Gynecol. 1999;181:1018- 355:120.
Weerdt J. Two rhombencephalic anencephalics: a 1025. 76. Myers LB, Bulich LA, Hess P, Miller NM. Fetal en-
clinico-pathological and electroencephalographic study. 53. Woo JS, Liang ST, Lo RL, Chan FY. Middle cere- doscopic surgery: indications and anaesthetic
Brain. 1968;91:497-506. bral artery Doppler flow velocity waveforms. Obstet management. Best Pract Res Clin Anaesthesiol. 2004;
33. Fisch BJ, Spehlmann R. Fisch and Spehlmann’s EEG Gynecol. 1987;70:613-616. 18:231-258.
Primer: Basic Principles of Digital and Analog EEG. 54. Wladimiroff JW, vd Wijngaard JA, Degani S, Noor- 77. Myers LB, Cohen D, Galinkin J, Gaiser R, Kurth
3rd ed. New York, NY: Elsevier; 1999. dam MJ, van Eyck J, Tonge HM. Cerebral and um- CD. Anaesthesia for fetal surgery. Paediatr Anaesth.
34. Scher MS. Electroencephalography of the new- bilical arterial blood flow velocity waveforms in nor- 2002;12:569-578.
born: normal and abnormal features. In: Nieder- mal and growth-retarded pregnancies. Obstet Gynecol. 78. Gregory GA, Wade JG, Beihl DR, Ong BY, Sitar
meyer E, Lopes da Silva FH, eds. Electroencephalog- 1987;69:705-709. DS. Fetal anesthetic requirement (MAC) for halothane.
raphy: Basic Principles, Clinical Applications, and 55. Giannakoulopoulos X, Sepulveda W, Kourtis P, Anesth Analg. 1983;62:9-14.
Related Fields. 4th ed. Philadelphia, Pa: Lippincott Wil- Glover V, Fisk NM. Fetal plasma cortisol and beta- 79. Gaiser RR, Kurth CD. Anesthetic considerations
liams & Wilkins; 1999:896-946. endorphin response to intrauterine needling. Lancet. for fetal surgery. Semin Perinatol. 1999;23:507-
35. Sharbrough FW. Nonspecific abnormal EEG 1994;344:77-81. 514.
©2005 American Medical Association. All rights reserved. (Reprinted) JAMA, August 24/31, 2005—Vol 294, No. 8 953
Downloaded from www.jama.com by guest on February 20, 2010

9. FETAL PAIN
80. Clyburn PA. Early thoughts on “Why Mothers Die 87. Fisk NM, Gitau R, Teixeira JM, Giannakoulopou- 93. Missant C, Van Schoubroeck D, Deprest J,
2000-2002.” Anaesthesia. 2004;59:1157-1159. los X, Cameron AD, Glover VA. Effect of direct fetal Devlieger R, Teunkens A, Van de Velde M. Remi-
81. Munnur U, Suresh MS. Airway problems in opioid analgesia on fetal hormonal and hemody- fentanil for foetal immobilisation and maternal
pregnancy. Crit Care Clin. 2004;20:617-642. namic stress response to intrauterine needling. sedation during endoscopic treatment of twin-to-
82. Royal College of Obstetricians and Gynaecologists. Anesthesiology. 2001;95:828-835. twin transfusion syndrome: a preliminary dose-
Confidential Enquiry Into Maternal and Child Health: 88. Loftus JR, Hill H, Cohen SE. Placental transfer and finding study. Acta Anaesthesiol Belg. 2004;55:
Why Mothers Die 2000-2002: Report on Confiden- neonatal effects of epidural sufentanil and fentanyl ad- 239-244.
tial Enquiries Into Maternal Deaths in the United ministered with bupivacaine during labor. 94. Van de Velde M, Van Schoubroeck D, Lewi LE,
Kingdom. London, England: RCOG Press; 2004. Anesthesiology. 1995;83:300-308. et al. Remifentanil for fetal immobilization and ma-
83. Atrash HK, Cheek TG, Hogue CJ. Legal abortion 89. Rayburn W, Rathke A, Leuschen MP, Chleborad ternal sedation during fetoscopic surgery: a random-
mortality and general anesthesia. Am J Obstet Gynecol. J, Weidner W. Fentanyl citrate analgesia during labor. ized, double blind comparison with diazepam. Anesth
1988;158:420-424. Am J Obstet Gynecol. 1989;161:202-206. Analg. 2005;101:251-258.
84. Atrash HK, MacKay HT, Binkin NJ, Hogue CJ. Le- 90. Jauniaux E, Jurkovic D, Lees C, Campbell S, Gul- 95. Lehmann KA, Gerhard A, Horrichs-Haermeyer G,
gal abortion mortality in the United States: 1972 to bis B. In-vivo study of diazepam transfer across the Grond S, Zech D. Postoperative patient-controlled an-
1982. Am J Obstet Gynecol. 1987;156:605-612. first trimester human placenta. Hum Reprod. 1996;11: algesia with sufentanil: analgesic efficacy and mini-
85. Bartlett LA, Berg CJ, Shulman HB, et al. Risk fac- 889-892. mum effective concentrations. Acta Anaesthesiol
tors for legal induced abortion-related mortality in the 91. Haram K, Bakke OM. Diazepam as an induction Scand. 1991;35:221-226.
United States. Obstet Gynecol. 2004;103:729- agent for caesarean section: a clinical and pharmaco- 96. Strumper D, Durieux ME, Gogarten W, Van Aken
737. kinetic study of fetal drug exposure. Br J Obstet H, Hartleb K, Marcus MA. Fetal plasma concentra-
86. Henshaw SK, Finer LB. The accessibility of abor- Gynaecol. 1980;87:506-512. tions after intraamniotic sufentanil in chronically in-
tion services in the United States, 2001. Perspect Sex 92. Gerdin E, Rane A, Lindberg B. Transplacental trans- strumented pregnant sheep. Anesthesiology. 2003;98:
Reprod Health. 2003;35:16-24. fer of morphine in man. J Perinat Med. 1990;18:305-312. 1400-1406.
The polymorphic visions of the eyes and the spirit are
contained in the uniform lines of small or capital let-
ters, periods, commas, parentheses—pages of signs,
packed as closely together as grains of sand, repre-
senting the many-colored spectacle of the world on a
surface that is always the same and always different,
like dunes shifted by the desert wind.
—Italo Calvino (1923-1985)
954 JAMA, August 24/31, 2005—Vol 294, No. 8 (Reprinted) ©2005 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on February 20, 2010