19
An fMRI Investigation of Moral Cognition in Healthcare Decision Making Lisa J. Smith Flinders University Paul Anand The Open University and Oxford University Abdelmalek Benattayallah University of Exeter Timothy L. Hodgson University of Lincoln This study used fMRI to investigate the neural substrates of moral cognition in health resource allocation decision problems. In particular, it investigated the cognitive and emotional processes that underpin utilitarian approaches to health care rationing such as Quality Adjusted Life Years (QALYs). Participants viewed hypothetical medical and nonmedical resource allocation scenarios which described equal or unequal allocation of resources to different groups. In addition, partici- pants were assigned to 1 of 2 treatments in which they either did or did not receive advanced instructions about the principles of utilitarianism. In all cases, partici- pants were asked to judged the proposed allocations as “fair” or “unfair.” More brain activity was observed within the superior parietal lobe, angular gyrus, middle temporal gyrus, and bilateral caudate nucleus when participants viewed scenarios depicting equal divisions of resources. Conversely, unequal resource divisions were associated with more activity in the inferior frontal gyrus and insula cortex. Furthermore, instructions about the principles of utilitarianism led to significant activation differences within the inferior frontal gyrus and the middle frontal gyrus. Significant differences in activity were also found within the inferior frontal cortex and anterior insula between medical and nonmedical scenarios. The implications for cognitive control mechanisms and the cognitive and neural bases of utilitarian ethical judgment are discussed. Keywords: equality, fMRI, health care, QALY, utilitarianism How to divide limited public resources for health care fairly among a population is a major moral and political issue in all countries around the world. Research has shown that people are attentive to perceived breaches of individual rights to treatment when asked to make health care rationing decisions for themselves (Anand & Walloo, 2000). In contrast, utilitarian per- spectives argue that the ‘rational’ way in health care resource allocation is not to heavily weight universal rights, but to allocate resources which maximize the widest public good. Neuroscientific research into moral cognition has emphasized the importance of both emotion and cognition in forming choices, although to date techniques of neuroscience and neuropsy- chology have not been applied to understanding health care rationing decision making (Ciara- melli et al., 2007; Koenigs et al., 2007; Moll, de Oliveira-Souza & Eslinger, 2003; Moll, Es- linger & de Oliveira-Souza, 2001; Moll et al., 2002; Moll & de Oliveira-Souza, 2007; Young & Koenigs, 2007; Koenigs et al., 2007). As well as being of interest in its own right, the question Lisa J. Smith, School of Psychology, Flinders University; Paul Anand, Department of Economics, Faculty of Social Sciences, The Open University, and Health Economics Re- search Centre, Nuffield Department of Public Health, Ox- ford University; Abdelmalek Benattayallah, Exeter Mag- netic Resonance Imaging Research Centre, University of Exeter; Timothy L. Hodgson, School of Psychology, Uni- versity of Lincoln. Correspondence concerning this article should be ad- dressed to Timothy L. Hodgson, School of Psychology, University of Lincoln, Brayford Campus, Lincoln LN6 7TS, United Kingdom. E-mail: [email protected] This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Journal of Neuroscience, Psychology, and Economics © 2015 American Psychological Association 2015, Vol. 8, No. 2, 116 –133 1937-321X/15/$12.00 http://dx.doi.org/10.1037/npe0000038 116 tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S1 6/7/15 23:19 Art: 2013-1039 APA NLM

An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

An fMRI Investigation of Moral Cognitionin Healthcare Decision Making

Lisa J. SmithFlinders University

Paul AnandThe Open University and Oxford University

Abdelmalek BenattayallahUniversity of Exeter

Timothy L. HodgsonUniversity of Lincoln

This study used fMRI to investigate the neural substrates of moral cognition inhealth resource allocation decision problems. In particular, it investigated thecognitive and emotional processes that underpin utilitarian approaches to healthcare rationing such as Quality Adjusted Life Years (QALYs). Participants viewedhypothetical medical and nonmedical resource allocation scenarios which describedequal or unequal allocation of resources to different groups. In addition, partici-pants were assigned to 1 of 2 treatments in which they either did or did not receiveadvanced instructions about the principles of utilitarianism. In all cases, partici-pants were asked to judged the proposed allocations as “fair” or “unfair.” Morebrain activity was observed within the superior parietal lobe, angular gyrus, middletemporal gyrus, and bilateral caudate nucleus when participants viewed scenariosdepicting equal divisions of resources. Conversely, unequal resource divisions wereassociated with more activity in the inferior frontal gyrus and insula cortex.Furthermore, instructions about the principles of utilitarianism led to significantactivation differences within the inferior frontal gyrus and the middle frontal gyrus.Significant differences in activity were also found within the inferior frontal cortexand anterior insula between medical and nonmedical scenarios. The implicationsfor cognitive control mechanisms and the cognitive and neural bases of utilitarianethical judgment are discussed.

Keywords: equality, fMRI, health care, QALY, utilitarianism

How to divide limited public resources forhealth care fairly among a population is a majormoral and political issue in all countries aroundthe world. Research has shown that people areattentive to perceived breaches of individualrights to treatment when asked to make health

care rationing decisions for themselves (Anand& Walloo, 2000). In contrast, utilitarian per-spectives argue that the ‘rational’ way in healthcare resource allocation is not to heavily weightuniversal rights, but to allocate resources whichmaximize the widest public good.

Neuroscientific research into moral cognitionhas emphasized the importance of both emotionand cognition in forming choices, although todate techniques of neuroscience and neuropsy-chology have not been applied to understandinghealth care rationing decision making (Ciara-melli et al., 2007; Koenigs et al., 2007; Moll, deOliveira-Souza & Eslinger, 2003; Moll, Es-linger & de Oliveira-Souza, 2001; Moll et al.,2002; Moll & de Oliveira-Souza, 2007; Young& Koenigs, 2007; Koenigs et al., 2007). As wellas being of interest in its own right, the question

Lisa J. Smith, School of Psychology, Flinders University;Paul Anand, Department of Economics, Faculty of SocialSciences, The Open University, and Health Economics Re-search Centre, Nuffield Department of Public Health, Ox-ford University; Abdelmalek Benattayallah, Exeter Mag-netic Resonance Imaging Research Centre, University ofExeter; Timothy L. Hodgson, School of Psychology, Uni-versity of Lincoln.

Correspondence concerning this article should be ad-dressed to Timothy L. Hodgson, School of Psychology,University of Lincoln, Brayford Campus, Lincoln LN6 7TS,United Kingdom. E-mail: [email protected]

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

Journal of Neuroscience, Psychology, and Economics © 2015 American Psychological Association2015, Vol. 8, No. 2, 116–133 1937-321X/15/$12.00 http://dx.doi.org/10.1037/npe0000038

116

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 2: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

of the extent to which brain areas involved inemotion and cognitive control contribute tosuch decision making might potentially informthe reason why people make the choices they doin these context. In the present study, we usefunctional MRI (fMRI) to examine how brainregions involved in decision making, cognition,and emotion are recruited in health care ration-ing decision making.

Utilitarianism, or more specifically ‘Act Util-itarianism,’ is a moral framework which pro-poses that one should choose actions that yieldat least as high a utility as the sum total ofutilities of any other alternative act (Bentham,1789). This moral framework has been highlyinfluential within economics and has given riseto the use of a QALY (Quality Adjusted LifeYear) maximization rule by economists work-ing on issues to do with health. Quality of lifefactors include loss of function and mobilityand the enjoyment one can derive from lifeafter resources have been allocated (Cubbon,1991). In essence, the expected QALY benefitof a medical intervention is the measure of itsaverage value and the set of interventionswhich should be funded according to utilitar-ianism is that which maximizes the total gainin QALYs by a society (Appleby, Devlin, &Parkin, 2007).

However, utilitarian resource allocation hasbeen criticized by philosophers concernedabout justice and distributional aspects offairness within a “Rawlsian” perspective(Rawls, 1971). According to Rawls each per-son has an equal right to the most extensivebasic liberty (greatest liberty principle) andinequalities are only permitted if they benefitall persons. For something to be just or fair itmust not violate principles of liberty or equal-ity. Utilitarianism permits distributional in-equalities if they serve the general interest incontrast to the egalitarian stance of justiceoutlined by Rawls.

QALY maximization is perceived by some ascondemning people who are not ‘cost-effective’to society if their care does not produce enoughmedical benefit. If all other things are equal, theQALY approach could prioritize the health ofsomeone who already has a high health status.A person who is disabled will be viewed ashaving a poorer long term quality of life and soany QALY score they obtain for treatment willbe inevitably lower than that of a healthy per-

son. Some philosophers (e.g., Harris, 1987,1995, 2005) object to the use of QALYs on thisbasis, arguing that it puts such chronically ill ordisabled persons in a sort of ‘double jeopardy.’Not only do they suffer a misfortune by becom-ing disabled or otherwise disadvantaged in thefirst instance, they are as a result given a lowerpriority in their health care. Empirical researchalso demonstrates that people do not make de-cisions consistent with QALYs (Anand & Wal-loo, 2000). It has been suggested that the QALYapproach fails to account for real choice behav-ior because it does not reflect the moral beliefsheld by ordinary members of the public whooften see the results of QALY decision makingframeworks as violating claims based on humanrights or social contracts (Anand & Walloo,2000).

In contrast to moral decision-making frame-works such as utilitarianism and QALYs whichemphasizes rational calculation, research incognitive neuroscience emphasizes the contri-bution of emotion to decision making in realsocial situations (e.g., Bechara & Damasio,2005). Analysis of brain activity via fMRI dur-ing economic “games” and moral decision mak-ing tasks reveals interacting networks of activa-tion within prefrontal and subcortical structuresassociated with value formation, cognition andemotion. For example, Sanfey, Rilling, Aron-son, Nystrom, and Cohen (2003) used fMRI tomeasure brain activity in the ‘ultimatum game,’in which two players are given the opportunityto split a sum of money: one player acts as theproposer and the responder must either acceptor reject the offer. Traditional economic theorywould predict that the proposer should offer thesmallest sum of money possible and that theresponder should accept on the premise thatsome money is preferable to none. The bilateralanterior insula, dorsolateral prefrontal cortex(DLPFC), and anterior cingulate cortex gener-ate greater activation in responder’s brains forunfair compared with fair offers. The insula inparticular has been associated with the experi-ence of disgust, an emotion that has been con-jectured to play an important role in moral de-cision making (Haidt, Rozin, McCauley, &Imada, 1997; Moll et al., 2005).

Greene, Nystrom, Engell, Darley, and Cohen(2004) used fMRI to study variants of the clas-sic “trolley car” problem. In the original versionof this problem, participants were asked to

117FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 3: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

make a hypothetical decision to save 5 lives ina runaway trolley car at the cost of sacrificingone individual by pulling a lever to divert thetrain onto an alternative track. In the “bridgevariant,” the 5 lives can only be saved by push-ing a fat man off a bridge into the path of thetrain. Green et al. found that brain areas asso-ciated with emotion and cognition exhibitedhigher levels of activation for personal moraljudgments (e.g., the bridge variant problem)versus impersonal moral judgments. More dif-ficult personal moral judgments showed in-creasing activity in the DLPFC and the ACC.By contrast, increased activity was found in theDLPFC only for impersonal moral judgmentswhich resembled utilitarian decisions.

Based on these findings, Green and col-leagues argued that difficult personal moraljudgments involve a conflict between cognitiveand emotional processes. They proposed a ‘dualprocess’ model of moral cognition in whichpersonal moral judgments are driven by socialemotional processes and impersonal moraljudgments are more cognitive in nature.

Subsequent work has also suggested that util-itarian moral judgments are driven by con-trolled cognitive processes, whereas nonutilitar-ian judgments are driven by automaticemotional responses (Greene, 2007). Greene,Morelli, Lowenberg, Nystrom, and Cohen(2008) presented participants with personal andimpersonal moral dilemmas either under cogni-tive load (digit search task) or a control condi-tion (no cognitive load). They found that utili-tarian judgments were significantly slower thannonutilitarian judgments under cognitive load.Greene and colleagues argue that this providesdirect evidence for their view that utilitarianjudgments are driven by controlled cognitiveprocesses, whereas nonutilitarian judgments aredriven by automatic (more emotional/affective)reasoning processes. In other research, it hasbeen found that patients with lesions of theventro-medial prefrontal cortex (VMPFC) pro-duced an abnormally high rate of utilitarianjudgments in cases which required the individ-ual to overcome an emotional response to thedirect harming of others (Koenigs et al., 2007).These patients also exhibited more normal judg-ments for impersonal situations, suggesting thatthe VMPFC is a crucial brain structure for themediation of moral judgments that involveemotionally salient actions.

Based on these previous investigations ofnon– health care decision making scenarios,we hypothesized that when people form deci-sions based on QALY maximization princi-ples, increased engagement of neural centersresponsible for cognitive control and reason-ing should be observed (e.g., DLPFC). Ourrationale is that when individuals make a util-itarian judgment, it is likely to require thesuppression of an immediate and powerfulemotional response to the perceived inter-group inequality and breach of human rightsto health care. Conversely, it is hypothesizedthat when people make decisions which goagainst QALY maximization, and choose notto allocate resources to those with greaterexpected utilities, then this will be associatedwith enhanced neural responses associatedwith emotions (e.g., anterior insula). More-over, as health care rationing might be consid-ered a particularly personal moral dilemma, en-hanced activity in emotional centers and moreRawlsian or rights-based decision makingshould be seen in such problems. This studyinvestigated these hypotheses using fMRI scan-ning while participants considered a series ofhypothetical resource allocation split scenarios.In each scenario, they indicated with a buttonpress whether they considered them to be “fair”or “unfair.”

Method

Participants

Thirty participants were recruited from ad-vertisements placed within the University ofExeter campus and from internal mailing lists.Participants were paid 10 pounds and were agedbetween 19 and 36 years (M � 26; 5 male, 25female). Ethical approval was obtained from theUniversity of Exeter School of Psychology eth-ics committee. Exclusion criteria specified weremajor medical illness (e.g., congenital abnor-malities, heart problems, cancer) neurologicaldisorders (e.g., Seizures, Head Injury), or metalfragments/implants which make MRI contrain-dicated.

Half the participants received instructionsabout utilitarian/QALY decision making frame-works as part of the information and consent

118 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 4: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

procedure1 (QALY Instructed participants). Theother group received no advanced briefing onutilitarian decision making (Uninstructedgroup).

Task and Materials

The experimental task involved the presenta-tion of 60 stimulus slides which required par-ticipants to make a button press response toindicate whether they considered a resource al-location/‘split’ of money between two socialgroups (defined by QALY relevant factors) tobe either fair or unfair. A control task involvedjudgments about how correctly (or “fairly”) ahorizontal bar indicated a stated quantitativesplit of a commodity. Stimuli used in the fMRIstudy were designed using Microsoft Power-Point and then exported into “jpg” format at1024 � 768 pixel resolution. Each scenarioscreen was displayed for up to 20 seconds,followed by presentation of a fixation cross fora variable period of 3 to 12 seconds. Partici-pants viewed the stimuli in the scanner via amirror mounted on the head coil and a backprojection screen located at the foot of the scan-ner. Responses were made using a fiber opticresponse key box. Custom written C-softwarewas used to display the image, record responsesand control stimulus timing via a laptop PC inthe scanner control room.

A horizontal bar in the stimulus screen wasdesigned to graphically represent the hypothet-ical allocation of resources (always a split of£10 million) to two groups who differed onQALY relevant factors (i.e., married versus nonmarried, children versus no children, high ver-sus low income, and old versus young). Thestimulus set of 60 slides contained, for everyparticipant, 20 Medical scenarios, 20 Nonmed-ical scenarios, and 20 Control scenarios. TheMedical scenarios were organized into twogroups: Group A (depression, diabetes, liverdisease, dementia, cancer) and Group B (HIV,asthma, epilepsy, influenza, heart disease). TheNonmedical scenarios comprised Group A (bas-ketball lessons, super high speed internet, uni-versity tuition fees, housing benefit and cyclingproficiency lessons) and Group B (driving les-sons, computer tuition, cookery lessons, road-side assistance and public transport passes) withparticipants viewing scenarios selected fromonly one Medical and one Nonmedical group in

each cases. The control scenarios were alwaysstatic across participants: quantity of chocolateversus carrot cakes, weight of two parcels, goalsscored by a football team, size of two officebuildings, number of students in two lectureclasses.

Content for Medical scenarios was based onsimilar scenarios used in Anand and Walloo(2000). Content for Nonmedical scenarios wereselected to include interventions which added tooverall quality of life while not directly impact-ing on health and longevity (e.g., having a fulldriving license leads to increased social mobil-ity). The scenarios were also subject to a pilotstudy in which participants were asked to spec-ify the division of resources they viewed as fairfor a variety of scenarios. Control scenarioswere chosen to reflect familiar everyday con-structs not related to moral judgments in anyway. Examples of screens presented to partici-pants in the scanner are shown in Figure 1.

Each slide differed in the amount of resourcesallocated to the groups and resource allocationwas either (a) in favor of ‘high’ QALY groups(70/30 split or 60/40 split), (b) was equal (50/50), or (c) in favor of ‘low’ QALY groups(40/60, 30/70). Participants were told that thenumber of people falling into each category(e.g., old v young) was equal. Decisions in favorof ‘high’ QALY groups were therefore consid-ered ‘utilitarian’ and decisions in favor of equalsplits or ‘low’ QALY groups were considered‘nonutilitarian.’

During the task the images displayed toparticipants were selected from one of fivestimulus sets. The sets differed in that thesame split (i.e., 70/30) was never combinedwith the same scenario (i.e., cancer treatment)

1 Instruction text read by these participants was as fol-lows: “Utilitarianism is the philosophical perspective thatmoral/ethical decisions should be made with the aim ofmaximizing the greatest good or ‘utility’ to the greatestnumber of people. This approach is often applied to deci-sions in health care provision such that those that are likelyto gain the most from, for example, a new drug treatmentare given priority. Calculations can be based on so calledquality adjusted life years (QALYs), which calculate thelikely life expectancy of an individual multiplied by theirquality of life. For example, someone who is married shouldbe prioritized for treatment as on average they have a betterquality of life and live longer. Similarly, those who havechildren, younger people and people who are financiallybetter off would have a higher QALY score than thosewithout children, older people or less well off individuals.”

119FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

Fn1

F1

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 5: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

across sets. Within these sets the presentationof the stimulus slides were randomized acrossparticipants.

Design and Procedure

A 2 � 5 � 2 mixed factorial design wasemployed. The repeated measures independentvariables were Distribution of resources or

‘splits’ which had five levels, 70/30, 60/40, 50/50, 40/60, 30/70, and Scenario type, which hadtwo levels, Medical or Nonmedical, with a be-tween-group factor of Participant Instructions(whether or not instructions about QALY max-imization were given to participants). The de-pendent variables measured were ethical stand-point (i.e., was the fair allocation judged to be

Figure 1. Examples of Control, Medical, and Nonmedical choice scenarios presented toparticipants. Three QALY relevant medical scenarios are presented in the right panel of thefigure. Two QALY relevant Nonmedical scenarios are presented in the top two left panels ofthe figure. An example of a control scenario is presented in the bottom left panel. See theonline article for the color version of this figure.

120 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

O CN OL LI ON RE

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 6: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

the utilitarian distribution or the nonutilitariandistribution) and the pattern of neural activityobserved in the brain via the measurement offMRI/Blood Oxygenation Level Dependent(BOLD) signal response.

Informed consent was sought from each par-ticipant who also completed a safety checklistand received a set of written instructions for thetask and screen shot of an example scenario.Participants lay face up in the scanner withcushions placed around their head to improvestability. Participants held a two-button re-sponse pad in their right hand and were in-structed to press the left button if they felt theallocation of resources in each scenario was‘fair’ or the right button if they felt it was‘unfair.’ Between each screen a fixation crosswas presented. After the completion of thefMRI session, participants read a debrief formexplaining the rationale and background to thestudy.

fMRI Data Acquisition

A total of 455 Whole Brain 3D Echo Planarimages were acquired as the participant per-formed the task using the Phillips 1.5T Gy-roscan whole body Intera scanner at the ExeterMRI Research Centre, University of Exeter,U.K. A T2�-weighted echo planar sequence(EPI) was used with the following parameters:TR � 3000ms, TE � 45ms, flip angle � 90degrees, 35 transverse slices, resolution � 3mm � 3 mm � 3 mm. FOV � 240 mm as-cending acquisition.

fMRI Preprocessing and Analysis

Spatial preprocessing and analysis were per-formed using SPM-8 (http://www.fil.ion.ucl.ac.uk/spm/). Preprocessing of the data includedslice-timing correction (resliced to mean image,interpolation 4th degree B spline), volume reg-istration, realignment to first scan image (the 6realignment parameter variables being added asnuisance covariates in the 1st level statisticalmodel), normalization to EPI template image,smoothing (using a 6-mm FWHM kernel, i.e.,twice the voxel dimensions), and high pass fil-tering (SPM default of 128 second cut-off pe-riod used).

Each participant’s fMRI data was analyzedusing a statistical model comprising a series ofregressors derived by convolving task event on-

sets and durations with a canonical hemody-namic response function. The GLM approachwas used to assess correlations between themeasured BOLD response and regressors ofinterest. The 1st level statistical model includedregressors corresponding to the onset of thechoice screen for different resource splits forMedical, Nonmedical, and Control scenarios,such that for each subject, 12 different eventregressors were represented in the first levelstatistical model corresponding to the follow-ing:

Medical Scenarios with 70/30 resource split,Medical-60/40, Medical-50/50, Medical-40/60,Medical-30/70 Nonmedical-70/30, Nonmedical-60/40, Nonmedical-50/50, Nonmedical-40/60,Nonmedical-30/70, Control Fair Representa-tion of Split, Control Unfair RepresentationSplit.

One-sample t tests on the statistical regres-sors of interest were generated for each partic-ipant and entered in the 2nd level (“randomeffects”) analysis to look for statistically signif-icant patterns of activation consistent acrossparticipants. Decision period activity for partic-ipants was then analyzed using a 2 � 2 � 5factorial model with (Medical/Nonmedical),participant instruction (received/did not receivebriefing about QALYs) and split type (70/30,60/40, 50/50, 40/60, 30/70) as factors. Statisti-cal contrasts, comparing equal versus unequalsplits of resources were constructed by applyingpositive and negative t-contrast weightings tothe relevant regressors within the 2nd levelSPM model. Because of the previously reportedfinding of a bias toward judging equality as“fair” in health care rationing (Anand & Wal-loo, 2000; see also behavioral analysis below),the reported analysis focused on main and in-teraction effects of Equal versus Unequal splitscenarios, equating to the comparison betweenregressors representing trials where participantsviewed 50/50 split scenarios compared to allother resource splits (70/30, 60/40, 40/60,30,70). Parametric modeling of the differentlevels of trial split was not used because it wasconsidered that the key comparison of interestwas between equality and nonequality ratherthat the relative quantitative split across scenar-ios.

Two participants were excluded from the 2ndlevel fMRI factorial analysis described belowbecause of a large amount of missing behavioral

121FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 7: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

data where no button press response was given(one participant) and another for providing anidiosyncratic response pattern (judging all sce-narios as “fair”). Second level factorial analysiswas therefore based on results from 28 subjects.Results are reported at p � .001 threshold un-corrected for multiple comparisons.

Estimated Brodmann area (BA) referencenumbers and anatomical labels were generatedfor each reported activation focus by first con-verting MNI coordinates to Talaraich coordi-nates (Talairach & Tournoux, 1988) accordingto the “mni2tal” transformation (www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html). These co-ordinates were then entered into the TalaraichDaemon database client software (Lancaster etal., 2000) to generate relevant anatomical la-bels.

Results

Behavioral Data

Nonutilitarian versus utilitarian decisionmakers. Using a median split, participantswere classified as either ‘strong’ nonutilitarianor ‘weak’ nonutilitarian decision makers basedupon their responses to the scenarios (consistentwith previous behavioral investigations all sub-jects showed some bias toward nonutilitarian-ism as evidenced by a preference for objectiveequality when judging fairness). Analysis ofbehavioral choices showed that more subjectsheld a strong nonutilitarian “deontological” po-sition in the subject group who were not giveninformation about the application of QALYmaximization to resource allocation prior to theexperiment (10/15 subjects holding a strong an-tiutilitarian view when not given informationabout QALY versus 6/15 subjects who weregiven information about QALY). However, aPearson Chi Square showed that there was nosignificant relationship between informationabout QALYs given to participants during theconsent procedure and their ethical standpoint,�2(1), N � 28, p � .274.

Fairness decision behavior. The propor-tion of scenarios judged as “unfair” differedsignificantly across Medical and Nonmedicalscenarios, F(1, 27) � 5.72, p � .05, with par-ticipants judging significantly more Medicalscenarios as unfair (67%) than Nonmedical sce-narios (62%). Participants also judged signifi-

cantly more 50–50 or ‘equal splits’ as fair thanany of the utilitarian or nonutilitarian unequalsplits, F(4, 108) � 47.25, p � .001. The judg-ments of fairness differed across Medical andNonmedical scenarios according to the resourceallocation split, F(4, 108) � 7.86, p � .001.This interaction effect indicated that fairnesswas judged differently across medical and non-medical scenarios, F(1, 27) � 23.80, p � .001with participants showing a significantly greatertendency to judge unequal splits as unfair inMedical compared with Nonmedical scenarios(see Figure 2).

Reaction times. A mixed ANOVA (2 �5 � 2) with ethical view (Weak versus StrongNonutilitarian) as a between subjects factor wasused to analyze the relationship between ethicalview, allocation of resources (‘splits’), and timetaken to process the decision (RT). Participantswere classified into ‘weak’ or ‘strong’ nonutili-tarian positions using a median split based ontheir percentage of fair/unfair judgments to de-cisions with utilitarian weighted resource allo-cation (see above). No significant main or in-teraction effects were observed although therewas a trend toward a difference between partic-ipants RTs in Medical scenarios, F(1, 28) �3.48, p � .074. There was also no significantmain effect based upon ethical view, F(1, 25),p � .94, ns, or interaction between ethical viewand split (70–30, 60–40, 50–50, 40–6-, 30–70).

A further 2-way ANOVA directly comparedRTs for scenarios of equality with all thosedepicting unequal splits. For this analysis thedifference between Medical and Nonmedicalscenarios did reach significance, F(1, 28) �5.91, p � .021; Means for Medical � 8871 �312 ms and Nonmedical � 9359 � 345ms,but there was no difference between Equaland Unequal scenarios overall, F(1, 28) �0.39; Means for Equal � 8945 � 403 ms andUnequal: 9184 � 314 ms, or interaction effectbetween Scenario type and Split, F(1, 28) �0.20.

Finally, we also carried out an analysis ofRTs by subject choice (Fair/Unfair). A 2-wayANOVA with Choice and Scenario type as fac-tor showed no significant main effect of Choice,F(1, 27) � 1.55, or Scenario, F(1, 27) � 1.45,and no significant interaction between the twofactors, F(1, 27) � 1.51.

122 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

F2

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 8: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

fMRI Data

Equal versus unequal splits (main effect).Comparison of brain activity during presenta-tion of scenarios of equality—50/50 splits—versus inequality (all other splits) revealed in-creased activity for equal splits within theangular gyrus (BA39), parietal lobe (BA7),middle temporal gyrus (BA19), superior frontalgyrus (BA8), precentral gyrus (BA4), and thehead of caudate nucleus (Table 1; Figure 3A).The bilateral head of caudate locus also in-cluded some “activated” voxels overlappingthe adjacent frontal ventricle on the standardMNI template image (see Figure 4). These arelikely to be artifacts of the standard smooth-ing and normalization procedures used to pro-cess fMRI data (see discussion). No voxelswere activated during presentation of scenar-ios with unequal as compared to equal splitsoverall.

Medical versus Nonmedical decision mak-ing (main effect). Comparison of brain activ-ity between scenarios types (Medical/Nonmed-

ical scenarios) showed no significantlyactivated voxels.

Equality versus inequality in Medical ver-sus Nonmedical scenarios (interaction effect).The interaction effect comparing the effect ofequality (vs. inequality) in Medical and Non-medical scenarios revealed foci of significantlyincreased activation in the right inferior frontalgyrus (BA9/BA47), right insula (BA13), andinferior parietal lobe (BA40; exact coordinatesare given in Table 1 and Figure 3B). To furtherunderstand this interaction, we applied contrastweights to either the equal or unequal splitconditions separately, for both Medical andNonmedical problems. We also plotted the con-trast estimates for each of the component modelregressors in the interaction. Consistent with thedifferences observed in the behavioral data, thisrevealed that the source of the interaction effectappeared to be a larger difference in activitybetween Equal and Unequal splits for Medicalas compared with Nonmedical scenarios (Fig-ure 5A).

Figure 2. Proportion of scenarios judged fair versus unfair for the two scenario types(Medical/Nonmedical) plotted against resource allocation split. 70-30 splits represent thosejudged to be more utilitarian based on QALY factors, with 30-70 splits representing a bias infavor of groups who would be expected to have lower QALYs.

123FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

T1, F3

F4

F5

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 9: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

The comparison of activity when viewingtrials depicting equality with inequality inevita-bly compared experimental conditions whichcontained unequal trial numbers (as four ver-sions of each scenario presented different un-equal resource splits depicted scenarios whereasonly one version depicted equality). We there-fore repeated the interaction analysis utilizingonly trials depicting equality and those depict-ing the most “utilitarian” split of resources,which favored the group which might be seen tohave most to gain in terms of QALYs (e.g.,70:30 in favor of young relative to old people),thus ensuring an equal number of measurementscontributed to the analysis in each participant.This supplementary analysis revealed a nearidentical pattern of activity to that reported inthe main interaction analysis including the fociin the right inferior frontal gyrus and insulacortex.

Subject instruction (main effect). Compa-rison of brain activity associated with subjectinstruction (those who had received instructionsabout QALY maximization vs. those whom hadnot) showed no significantly activated voxels.

Equality and instruction (interactioneffect). The interaction effect comparing theeffect of equality (see above) for the two in-struction groups showed significant differencesin activation to the effect of scenario equalitydependent on whether or not participants re-ceived instructions concerning principles ofQALY maximization. Regions activated in-cluded the inferior frontal gyrus (BA47) andmiddle frontal gyrus (BA11). Exact coordinatesare given in Table 1 and illustrated in Figure3C. As with the interaction between scenariotype and equality (see above) we applied con-trast weights to either the Equal or Unequal splitconditions separately for Medical and Nonmed-

Table 1Table of Peak Activation Coordinates (MNI Coordinates) and Anatomical Labels for the StatisticalContrasts of fMRI/BOLD Signal Response Described in the Text

MNI coordinates

Contrast/region (Brodmann areas) Z score x y z

Main effect equal � unequal splitsAngular gyrus (BA39) 6.60 45 �58 10Superior parietal lobe (BA7) 6.52 �18 �79 34Middle temporal gyrus (BA19) 6.13 39 �79 16Bilateral head of caudate 5.20 6 11 13

5.20 �6 17 7Pre-central gyrus (BA4) 4.66 21 �25 70Superior frontal gyrus (BA8) 4.58 12 44 52Superior parietal lobule (BA7) 4.26 21 �55 55Pre-central gyrus (BA6) 3.85 �51 �4 46Superior temporal gyrus (BA41) 3.12 42 �31 16

Interaction effect between unequal � equal contrast for nonmedicalversus medical scenarios

Inferior frontal gyrus (BA9/47) 5.01 45 8 22Insula (BA13) 4.73 39 2 4Inferior parietal lobe (BA40) 4.09 60 �37 37

Interaction effect between unequal � equal and participant instructiongroup (participants briefed about QALYs or not)

Inferior frontal gyrus (BA47) 3.40 39 23 �17Middle frontal gyrus (BA11) 3.15 �30 38 �2

Comparison of activity between experimental versus control trials overallLimbic lobe/posterior cingulate (BA30) 5.72 �21 �55 10Caudate 5.75 �21 �37 16Parietal lobe (BA7) 5.81 �9 �49 52Post-central gyrus (BA47) 4.91 �48 �16 22Superior frontal gyrus (BA6) 4.47 �18 11 52Medial frontal gyrus (BA9) 3.93 27 41 43Pre-central gyrus (BA4) 3.66 48 �10 52

124 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 10: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

ical problems and plotted the contrast estimatesfor each of the component model regressors inthe interaction. This showed that activity in theinferior frontal gyrus and middle frontal gyruswas increased for scenarios of inequality in theInstructed subject group (Figure 5B). As for theinteraction effect by scenario type above, thisanalysis was repeated including only trials de-picting either equality or 70:30 splits in favor ofhigh QALY gain groups. This showed a similarlocus of activity in the right inferior frontal

gyrus as well additional loci within the leftinferior and anterior cingulate gyrus (BA32).

Comparison by ethical view. A separatestatistical model comparison of brain activityshowed no significant main or interaction effecton brain activation between those who wereclassified as Weak versus Strong Nonutilitariandecision makers (see Behavioral Data above).

Experimental versus control trial compa-rison. The main effect comparison of Exper-imental versus Control trials overall revealed

Figure 3. (A) Significant regions of activity for scenarios describing equal relative tounequal resource splits (main effect of equality; color bar shows t statistic for all images). (B)Increased activity observed in inferior frontal cortex/insula cortex region to unequal scenariosfor Nonmedical relative to Medical scenarios (Equality by Scenario type interaction effect).(C) Area of inferior frontal cortex showing increased activation when viewing unequalscenarios (Medical and Nonmedical) for participants given instructions explaining principlesof QALY/Utilitarian decision making relative to those who were not (Interaction effectbetween equality and Instruction group). Color Bars indicate t statistic for each contrastdisplayed. See the online article for the color version of this figure.

125FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

O CN OL LI ON RE

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 11: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

widespread activation for both comparisons ofscenarios depicting equal and unequal splits ofresources or commodities, including enhancedactivity for the experimental condition in re-gions including the limbic lobe and posteriorcingulate (BA30), parietal lobe (BA7), temporallobe, and frontal lobe (BA6, BA4 and BA9; seeTable 1).

Discussion

Key Findings

The key finding of the current study was thatunder conditions for which participants moreoften judged unequal splits of resources as“fair,” relative increases in activity were seen inbrain areas linked to cognitive control processesincluding the DLPFC, inferior frontal cortex(IFC), and anterior cingulate gyrus. Specifi-cally, a greater difference in activation betweenunequal and equal splits of resources was seenin the right IFC and DLPFC (BA47 & BA9) formedical scenarios, compared with nonmedicalones. Inspection of the contrast estimates forthis interaction effect suggests a relative deac-tivation in these regions during viewing of med-

ical scenarios depicting uneven allocation ofresources to different patient groups, but nodifference in BOLD signal in this region depen-dent on resource split for Nonmedical scenarios(Figure 5A).

Implications for Accounts of MoralDecision Making

One interpretation of these results is thatmaking utilitarian judgments recruits brain re-gions in the frontal cortex associated with cog-nitive control functions and that there may bemore sustained engagement of these regionsduring consideration of Nonmedical comparedwith Medical scenarios and for participants whoreceived prior instruction about QALY baseddecision making. An explanation for this ob-served interaction effect might be that cognitivecontrol systems are disrupted by presentation ofscenarios perceived to breach ethical and socialnorms of equality of access to health care. Thisis consistent with the accompanying behavioraldata, which replicated the previously reportedbias for individuals to judge only equal resourcesplits as fair, particularly in medical scenarios.In these contexts, cognitive control may be lessengaged and participants might be more likely,as a result, to judge unequal allocations of re-sources as unfair.

The right IFC region highlighted in the presentstudy has been strongly implicated in cognitivecontrol processes. The role of the IFC in cognitivecontrol has been investigated predominantly usingthe Go/No-Go paradigm to test response inhibi-tion where the index of inhibitory control is thenumber of errors made when an individual makesa button press response to a trial they should not(Aron, Robbins, & Poldrack, 2004). The IFC be-comes activated during these trials of responseinhibition (Konishi et al., 1999; Liddle, Kiehl, &Smith, 2001; Menon, Adl3eman, White, Glover,& Reiss, 2001; Rubia, Smith, Brammer, & Taylor,2003). Moreover, damage to the right inferiorfrontal gyrus leads to disruption in stop-signalcued response inhibition (Aron, Fletcher, Bull-more, Sahakian, & Robbins, 2003). Maturation ofthis area is also related to response inhibition, witha significant relationship between speed of inhibi-tion and age (Tamm, Menon, & Reiss, 2002). Theinhibitory control functions of this region havealso been shown to extend across motor response

Figure 4. Zoomed axial and coronal views of a statisticalparametric map overlay image (left), alongside the samesectional views of a T1 MNI standardized brain imagewithout the overlay (right) for the contrast of scenariosdepicting Equality � Inequality. Voxels fulfilling p � .001uncorrected for multiple comparisons statistical thresholdare shown located in the grey matter of the head of caudatenucleus bilaterally. See the online article for the color ver-sion of this figure.

126 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

O CN OL LI ON RE

tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039APA NLM

Page 12: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

modalities (Hodgson et al., 2007; Aron et al.,2004).

The IFC has previously been shown to play arole in social cognition and has been implicatedin the ability to predict the intentions of others(Brunet, Sarfarti, Hardy-Bayle, & Decety,2000). Using TMS to induce temporary lesions,

it has also been shown that the left inferiorfrontal gyrus plays a role in social perception(Keuken et al., 2011). There are direct connec-tions between the IFC and the medial prefrontalcortex, an area involved in self-other represen-tations and has been suggested that this pathwayfacilitates integration of information necessary

Figure 5. Contrast estimate plots for peak activation voxels in the inferior frontal gyrus forthe interaction effect contrasts shown in Figure 3 and Table 1 (MNI coordinates shown).Vertical scale approximates to mean Beta coefficients fitted to the BOLD signal data for eachtrial/event regressor within the first level SPM analysis. (A) Contrast estimates for Medicaland Nonmedical scenarios depicting Equal and Unequal splits of resources. (B) Contrastestimates for participants who were instructed or uninstructed in QALY decision making.

127FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 13: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

for self-other representations (Uddin, Iacoboni,Lange, & Keenan, 2007). Further support forthe role of the IFC in social cognition has beenshown by the involvement of this region inmore removed forms of social cognition such asthe watching of movie clips depicting everydaysocial interactions (Iacoboni et al., 2004) orduring responses to incongruent cues in socialstimuli for example, press left when a faceshowed a rightward directed gaze (Schilbach etal., 2011).

The current results are only partially consis-tent with previously outlined “dual process”accounts of moral judgment and decision mak-ing (Greene, 2007). Earlier studies have empha-sized the anterior cingulate region and theDLPFC as implementing cognitive control overprepotent emotional response (Greene et al.,2004). The current results indicate that the rel-ative contribution of the IFC region to decisionmaking processes also needs to be considered insuch contexts. The behavioral bias of partici-pants to judge only objective equality as fairalso suggests that learned social rules and norms(i.e., “Rawlsian” ideas of fairness) may be im-portant in shaping peoples judgment of healthcare resource rationing as are emotional sys-tems. As well as suppressing a conflicting emo-tional response, cognitive control systems mayalso need to be engaged to suppress a neuralresponse to violations of prevalent social valuesand norms in such scenarios (Hodgson, Guala,Miller, Summers, 2012) to implement moreutilitarian decision making frameworks.

The activation observed during judgments offairness in the IFC region was centered onBA47 but also extended rostrally and dorsallyinto other lateral frontal areas including BA11and BA9 (i.e., DLPFC). Activation was alsoobserved in the adjacent anterior insula duringthe task for the comparison of Medical scenar-ios compared with Nonmedical scenarios. It isworth noting that the insula has been previouslyimplicated in social decision making and par-ticularly in mediating the emotion of social dis-gust (Sanfey et al., 2003; Phillips et al., 1997;Adolphs, Tranel, & Damasio, 2003; Calder,Keane, Manes, Antoun, & Young, 2000; Carr,Iacoboni, Dubeau, Mazziotta, & Lenzi, 2003;Wright, He, Shapira, Goodman, & Liu, 2004).Further work would be needed to characterizeparticipants’ emotional responses to confirmthat disgust plays a critical role in guiding par-

ticipants’ responses in the current experiment.An alternative explanation is that insula activityin this context represents empathetic pain re-sponse in participants’ accompanying the per-ceived social injustice depicted in the scenario(Gu et al., 2012). Nonetheless, the fMRI resultssuggest that there may be something specialabout the processing of medical scenarios de-picting unequal resource splits which modulatesactivity in the insula region.

Another region that showed differential acti-vation in response to equality versus inequalityin the present study was an anterior temporallobe region around the middle temporal gyrus(see Table 1). Other research has shown themiddle temporal gyrus to be activated in socialcontexts for example when subjects empathizewith victims of crime (Farrow et al., 2001).Other studies have implicated this region in therepresentation of abstract conceptual knowl-edge of social behaviors (Zahn et al., 2007), andit has been proposed to form part of a wide-spread neuronal network that supports socialcognition by providing access to social conceptsor rules (Ross & Olson, 2010).

Interestingly, other research has shown thatmoral choices and preferences are subject tovariation dependent upon the form and contextwithin which a question is presented (Tversky& Kahneman, 1981). Our findings indicate thatthe instructions given to participants about thebasis of utilitarian decision making and QALYscan influence their choices and modify brainactivity (Figure 3C). Nord (1995) has demon-strated “framing effects” in a person trade-offscenario in which social value is estimated byasking people to state the number of outcomesof one kind they consider equal in social valueto another kind. One direction for future re-search in the area would be to further investi-gate framing effects in health care resource al-location decision making within a neuroethicalperspective. Another interesting study whichcould further investigate the role of cognitivecontrol and inhibitory processing would be toprovide participants with a secondary cognitivetask while they perform similar judgments tothose used here. Previous work has indicatedthat providing participants with an additionalcognitive load during decision making may dis-rupt utilitarian decision making, leading toquicker response times and reduced cognitive

128 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 14: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

control over decision making (Greene et al.,2008).

Limitations and Criticisms

This is a pioneering but relatively small-scalefMRI study of moral choice in health care de-cision making. As such it is important to discussthe limitations and criticisms of the study designand analysis.

The choice of control task used can be criti-cized. We designed the control task to be a closematch with the experimental task in terms of thevisual properties of the screen and the motoraction to be given by the participant (left orright button press). Other aspects of the controltrials were designed to be very different fromthe experimental task, with participants beingrequired to make a judgment as to whether thebar indicator matched the written description ofa quantitative split of a commodity. In thisrespect, control trials lacked any demands onemotional and moral decision making andplaced minimal cognitive demands on the par-ticipant. We acknowledge that other forms ofcontrol task could have been used, and the oneused here may be considered to be too differentto the experimental condition to provide a use-ful comparison. However, it is important to notethat for trial/event related fMRI designs such asused here, the effective baseline for compari-sons of activity with experimental trials is brainactivity during null periods between the deci-sion screen onsets. Most of the contrasts ofinterest reported reflect direct comparisons be-tween experimental trial types which wereclosely matched other than in a single variableand were independent of activity elicited by thecontrol task.

Another aspect of the task which should benoted is that participants always indicated theirattribution of fairness or unfairness by pressingthe left or the right response key on a button boxheld in their dominant (right hand) and themapping of finger to choice was not counterbalanced across participants. It is conceivabletherefore that some aspects of the results mightarise from activation differences associatedwith the movement of the index and middlefinger rather than processing of the presentedscenario. However, no activation differenceswere found in the primary motor cortex for anyof the contrasts of fair versus unfair judgments

and the reported analysis of activity was fo-cused on the epoch before response executionrather than after the response. Further, the mostinteresting aspect of the results were interactioneffects which could not be simply explained byany confounding effect of preparatory motoractivity. Overall, therefore, it seems very un-likely that this minor feature of the design hadan effect on the results.

The analysis of fMRI data used the generallinear model (GLM) approach. Although this isthe most common statistical technique for ana-lyzing fMRI data, it has been criticized forbeing too rigid, with minor mis-modeling re-sulting in material loss of statistical power(Lindquist, 2008). These issues apply to mostfMRI work at present and call either for the useof larger samples or more advanced analysistechniques (see for instance a Bayesian ap-proach developed by Woolrich, Behrens, Beck-mann, Jenkinson, & Smith, 2004). However, aspecific statistical issue affecting interpretationof the present experiment is the low number oftrials presented for some conditions of interest(e.g., four medical scenarios depicting equalityper participant). The number of stimulus pre-sentations was limited to minimize the amountof time participants spent in the scanner and toensure that participants attended to the taskthroughout (total task length was 28 minutesand total time in scanner per subject was around40–45 minutes). Although it is more commonpractice to average fMRI data over more trials,there is no commonly accepted minimum num-ber of trials per condition or participants foreffective fMRI investigations, with some stud-ies having successfully measured the BOLDresponse to single trials in single subjects (Posseet al., 2001; Mechelli, Price, Henson, & Friston,2003). The minimum number of trials requiredto measure a given BOLD response depends onthe size of the effect to be measured, togetherwith the number of participants tested and theinherent variability of the measured signal (allof which factor together to determine the statis-tical power of a given study).

Neuroimaging also requires that individualbrain images are transformed into a standard-ized brain space such that data can be aver-aged across subjects with different brain di-mensions and morphology. Spatial smoothingof signals reduces artifactual signal noise butplaces further constraints on the spatial accu-

129FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 15: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

racy of the technique. These preprocessingtechniques can lead to potential errors in pre-cisely coregistering activity within specificanatomical substrates, particularly for smalland closely adjacent regions. This is the mostlikely explanation for the artifactual “activa-tion” of a small number of voxels in thefrontal ventricles adjacent to the caudate nu-clei in the present study (see Figure 4).

A different source of general concern aboutapplying fMRI to moral and ethical situations ishighlighted by Schleim and Schirmann (2011),who argue that the normative implications anduncertainties are too readily ignored or disre-garded by some when discussing neuroimagingresearch. They propose that the findings offMRI studies do not by themselves yield nor-mative implications, that is, prescribe how oneought to act. It should also be noted that there isa strong “reverse inference” problem underly-ing interpretation of neuroimaging in such con-texts. Even relatively well-evidenced claimsabout functions mediated by particular brainareas (e.g., IFC mediates cognitive inhibitorycontrol) are still subject to debate and discus-sion within cognitive neuroscience. Brain imag-ing data cannot therefore necessarily be viewedas a more objective/definitive a line of enquiryas other approaches. The neuro-ethical perspec-tive offered here cannot resolve the ongoingdebates between libertarians and egalitarian the-orists but offers an interesting framework withinwhich to think about the issues (Casebeer &Churchland, 2003; Braveman & Gruskin, 2003;Sandel, 2009; Friesen, 2001; Leary, 1994).

Finally, it should be noted that all of ourparticipants were young (aged between 16 and36), native English speakers studying or work-ing in the U.K. Therefore the behavioral deci-sions and accompanying neural activity in theseparticipants might be quite different in partici-pants with experience of different health caresystems, social group/identity characteristics,and cultural values.

Concluding Remarks

This is the first study to use fMRI to inves-tigate moral fairness decision making judg-ments in health care resource allocation scenar-ios. Consistent with previous work using purelybehavioral measures (Anand & Walloo, 2000),we found evidence for an egalitarian decision

making bias in medical treatment rationing sce-narios. Differences in brain activity in responseto presentation of equal versus unequal splits ofresources in medical and nonmedical scenarioswere observed which provide indicators of themechanism underlying the behavioral bias to-ward objective equality observed in participantsin judgments of fairness. A particular role forthe inferior frontal cortex (IFC) is suggested inoverriding the bias toward judging only objec-tively equal resource splits as fair, whereasheightened activation of the insula may indicatea reaction of disgust toward breaches of socialnorms or empathetic pain evoked by perceivedinequality in other social groups’ access tohealth care.

The current study should be seen as making asmall contribution to a question of very largeglobal socioeconomic importance. As such wehave taken care to list the methodological andanalytical limitations and criticisms of thestudy. However, as an initial investigation of thebiological mechanisms underlying human deci-sion making in this area we hope the work mayprompt new ways of thinking about fairness andinequality that could be value to economists andphilosophers alike.

References

Adolphs, R., Tranel, D., & Damasio, A. R. (2003).Dissociable neural systems for recognizing emo-tions. Brain and Cognition, 52, 61–69. http://dx.doi.org/10.1016/S0278-2626(03)00009-5

Anand, P., & Walloo, A. (2000). Utilities versus rightto publically provided goods: Arguments and evi-dence from health care rationing. Economica, 67,543–577. http://dx.doi.org/10.1111/1468-0335.00224

Appleby, J., Devlin, N., & Parkin, D. (2007). NICE’scost effectiveness threshold. British Medical Jour-nal, 335, 358–359. http://dx.doi.org/10.1136/bmj.39308.560069.BE

Aron, A. R., Fletcher, P. C., Bullmore, E. T., Saha-kian, B. J., & Robbins, T. W. (2003). Stop-signalinhibition disrupted by damage to right inferiorfrontal gyrus in humans. Nature Neuroscience, 6,115–116. http://dx.doi.org/10.1038/nn1003

Aron, A. R., Robbins, T. W., & Poldrack, R. A.(2004). Inhibition and the right inferior frontalcortex. Trends in Cognitive Sciences, 8, 170–177.http://dx.doi.org/10.1016/j.tics.2004.02.010

Bechara, A., & Damasio, A. R. (2005). The somaticmarker hypothesis: A neural theory of economicdecision. Games and Economic Behavior, 52,

130 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 16: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

336–372. http://dx.doi.org/10.1016/j.geb.2004.06.010

Bentham, J. (1789). An introduction to the principlesof morals and legislation. London, UK: T. Payne& Son.

Braveman, P., & Gruskin, S. (2003). Defining equityin health. Journal of Epidemiology and Commu-nity Health, 57, 254 –258. http://dx.doi.org/10.1136/jech.57.4.254

Brunet, E., Sarfati, Y., Hardy-Baylé, M. C., & De-cety, J. (2000). A PET investigation of the attribu-tion of intentions with a nonverbal task. NeuroIm-age, 11, 157–166. http://dx.doi.org/10.1006/nimg.1999.0525

Calder, A. J., Keane, J., Manes, F., Antoun, N., &Young, A. W. (2000). Impaired recognition andexperience of disgust following brain injury. Na-ture Neuroscience, 3, 1077–1078. http://dx.doi.org/10.1038/80586

Carr, L., Iacoboni, M., Dubeau, M. C., Mazziotta,J. C., & Lenzi, G. L. (2003). Neural mechanisms ofempathy in humans: A relay from neural systemsfor imitation to limbic areas. PNAS Proceedings ofthe National Academy of Sciences of the UnitedStates of America, 100, 5497–5502. http://dx.doi.org/10.1073/pnas.0935845100

Casebeer, D. W., & Churchland, P. S. (2003). Theneural mechanisms of moral cognition: A multi-ple-aspect approach to moral judgment and deci-sion-making. Biology and Philosophy, 18, 169–194. http://dx.doi.org/10.1023/A:1023380907603

Ciaramelli, E., Muccioli, M., Ladavas, E., & di Pel-legrino, G. (2007). Selective deficit in personalmoral judgment following damage to ventromedialprefrontal cortex. Social Cognitive and AffectiveNeuroscience, 2, 84–92. http://dx.doi.org/10.1093/scan/nsm001

Cubbon, J. (1991). The principle of QALY maximi-sation as the basis for allocating health care re-sources. Journal of Medical Ethics, 17, 181–184.http://dx.doi.org/10.1136/jme.17.4.181

Farrow, T. F., Zheng, Y., Wilkinson, I. D., Spence,S. A., Deakin, J. F., Tarrier, N., . . . Woodruff,P. W. (2001). Investigating the functional anatomyof empathy and forgiveness. Neuroreport, 12,2433–2438. http://dx.doi.org/10.1097/00001756-200108080-00029

Friesen, T. (2001). The right to health care. HealthLaw Journal, 9, 205–222.

Greene, J. D. (2007). Why are VMPFC patients moreutilitarian? A dual-process theory of moral judg-ment explains. Trends in Cognitive Sciences, 11,322–323. http://dx.doi.org/10.1016/j.tics.2007.06.004

Greene, J. D., Morelli, S. A., Lowenberg, K., Nys-trom, L. E., & Cohen, J. D. (2008). Cognitive loadselectively interferes with utilitarian moral judg-

ment. Cognition, 107, 1144–1154. http://dx.doi.org/10.1016/j.cognition.2007.11.004

Greene, J. D., Nystrom, L. E., Engell, A. D., Darley,J. M., & Cohen, J. D. (2004). The neural bases ofcognitive conflict and control in moral judgment.Neuron, 44, 389–400. http://dx.doi.org/10.1016/j.neuron.2004.09.027

Gu, X., Gao, Z., Wang, X., Liu, X., Knight, R. T.,Hof, P. R., & Fan, J. (2012). Anterior insularcortex is necessary for empathetic pain perception.Brain: A Journal of Neurology, 135, 2726–2735.http://dx.doi.org/10.1093/brain/aws199

Haidt, J., Rozin, P., McCauley, C., & Imada, S.(1997). Body, psyche, and culture: The relation-ship between disgust and morality. Psychologyand Developing Societies, 9, 107–131. http://dx.doi.org/10.1177/097133369700900105

Harris, J. (1987). QALYfying the value of life. Jour-nal of Medical Ethics, 13, 117–123. http://dx.doi.org/10.1136/jme.13.3.117

Harris, J. (1995). Double jeopardy and the veil ofignorance—A reply. Journal of Medical Ethics,21, 151–157. http://dx.doi.org/10.1136/jme.21.3.151

Harris, J. (2005). It’s not NICE to discriminate. Jour-nal of Medical Ethics, 31, 373–375. http://dx.doi.org/10.1136/jme.2005.012906

Hodgson, T., Chamberlain, M., Parris, B., James, M.,Gutowski, N., Husain, M., & Kennard, C. (2007).The role of the ventrolateral frontal cortex in in-hibitory oculomotor control. Brain: A Journal ofNeurology, 130, 1525–1537. http://dx.doi.org/10.1093/brain/awm064

Hodgson, T. L., Guala, F., Miller, T., & Summers, I.(2012). Limbic and prefrontal activity during con-formity and violation of norms in a coordinationgame. Journal of Neuroscience, Psychology, andEconomics, 5, 1–17. http://dx.doi.org/10.1037/a0026584

Iacoboni, M., Lieberman, M. D., Knowlton, B. J.,Molnar-Szakacs, I., Moritz, M., Throop, C. J., &Fiske, A. P. (2004). Watching social interactionsproduces dorsomedial prefrontal and medial pari-etal BOLD fMRI signal increases compared to aresting baseline. NeuroImage, 21, 1167–1173.http://dx.doi.org/10.1016/j.neuroimage.2003.11.013

Keuken, M. C., Hardie, A., Dorn, B. T., Dev, S.,Paulus, M. P., Jonas, K. J., . . . Pineda, J. A. (2011).The role of the left inferior frontal gyrus in socialperception: An rTMS study. Brain Research,1383, 196 –205. http://dx.doi.org/10.1016/j.brainres.2011.01.073

Koenigs, M., Young, L., Adolphs, R., Tranel, D.,Cushman, F., Hauser, M., & Damasio, A. (2007).Damage to the prefrontal cortex increases utilitar-ian moral judgements. Nature, 446, 908 –911.http://dx.doi.org/10.1038/nature05631

131FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 17: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

Konishi, S., Nakajima, K., Uchida, I., Kikyo, H.,Kameyama, M., & Miyashita, Y. (1999). Commoninhibitory mechanism in human inferior prefrontalcortex revealed by event-related functional MRI.Brain: A Journal of Neurology, 122, 981–991.http://dx.doi.org/10.1093/brain/122.5.981

Lancaster, J. L., Woldorff, M. G., Parsons, L. M., Liotti,M., Freitas, C. S., Rainey, L., . . . Fox, P. T. (2000).Automated Talairach atlas labels for functional brainmapping. Human Brain Mapping, 10, 120–131.http://dx.doi.org/10.1002/1097-0193(200007)10:3�120::AID-HBM30�3.0.CO;2-8

Leary, V. A. (1994). Defining the right to health care.In A. R. Chapman (Ed.), Health care reform: Ahuman rights approach. Washington, DC: George-town University Press.

Liddle, P. F., Kiehl, K. A., & Smith, A. M. (2001).Event-related fMRI study of response inhibition.Human Brain Mapping, 12, 100–109. http://dx.doi.org/10.1002/1097-0193(200102)12:2�100::AID-HBM1007�3.0.CO;2-6

Lindquist, M. A. (2008). The statistical analysis offMRI data. Statistical Science, 23, 439 – 464.http://dx.doi.org/10.1214/09-STS282

Mechelli, A., Price, C. J., Henson, R. N. A., &Friston, K. J. (2003). Estimating efficiency a pri-ori: A comparison of blocked and randomized de-signs. NeuroImage, 18, 798 – 805. http://dx.doi.org/10.1016/S1053-8119(02)00040-X

Menon, V., Adleman, N. E., White, C. D., Glover,G. H., & Reiss, A. L. (2001). Error-related brainactivation during a Go/NoGo response inhibitiontask. Human Brain Mapping, 12, 131–143. http://dx.doi.org/10.1002/1097-0193(200103)12:3�131::AID-HBM1010�3.0.CO;2-C

Moll, J., & de Oliveira-Souza, R. (2007). Moraljudgments, emotions and the utilitarian brain.Trends in Cognitive Sciences, 11, 319–321. http://dx.doi.org/10.1016/j.tics.2007.06.001

Moll, J., de Oliveira-Souza, R., & Eslinger, P. J.(2003). Morals and the human brain: A workingmodel. Neuroreport, 14, 299–305. http://dx.doi.org/10.1097/00001756-200303030-00001

Moll, J., de Oliveira-Souza, R., Eslinger, P. J., Bra-mati, I. E., Mourão-Miranda, J., Andreiuolo, P. A.,& Pessoa, L. (2002). The neural correlates ofmoral sensitivity: A functional magnetic resonanceimaging investigation of basic and moral emo-tions. The Journal of Neuroscience, 22, 2730–2736.

Moll, J., de Oliveira-Souza, R., Moll, F. T., Ignácio,F. A., Bramati, I. E., Caparelli-Dáquer, E. M., &Eslinger, P. J. (2005). The moral affiliations ofdisgust: A functional MRI study. Cognitive andBehavioral Neurology, 18, 68–78. http://dx.doi.org/10.1097/01.wnn.0000152236.46475.a7

Moll, J., Eslinger, P. J., & de Oliveira-Souza, R.(2001). Frontopolar and anterior temporal cortex ac-

tivation in a moral judgment task: Preliminary func-tional MRI results in normal subjects. Arquivos deNeuro-Psiquiatria, 59, 657– 664. http://dx.doi.org/10.1590/S0004-282X2001000500001

Nord, E. (1995). The person-trade-off approach tovaluing health care programs. Medical DecisionMaking, 15, 201–208. http://dx.doi.org/10.1177/0272989X9501500302

Phillips, M. L., Young, A. W., Senior, C., Brammer,M., Andrew, C., Calder, A. J., . . . David, A. S.(1997). A specific neural substrate for perceivingfacial expressions of disgust. Nature, 389, 495–498. http://dx.doi.org/10.1038/39051

Posse, S., Binkofski, F., Schneider, F., Gembris,D., Frings, W., Habel, U., . . . Eickermann, T.(2001). A new approach to measure single-eventrelated brain activity using real-time fMRI: Fea-sibility of sensory, motor, and higher cognitivetasks. Human Brain Mapping, 12, 25– 41. http://dx.doi.org/10.1002/1097-0193(200101)12:1�25::AID-HBM30�3.0.CO;2-H

Rawls, J. (1971). A theory of justice. Cambridge,MA: Harvard University Press.

Ross, L. A., & Olson, I. R. (2010). Social cognitionand the anterior temporal lobes. NeuroImage, 49,3452–3462. http://dx.doi.org/10.1016/j.neuroim-age.2009.11.012

Rubia, K., Smith, A. B., Brammer, M. J., & Taylor,E. (2003). Right inferior prefrontal cortex medi-ates response inhibition while mesial prefrontalcortex is responsible for error detection. Neuro-Image, 20, 351–358. http://dx.doi.org/10.1016/S1053-8119(03)00275-1

Sandel, M. (2009). Justice: What’s the right thing todo? New York, NY: Penguin.

Sanfey, A. G., Rilling, J. K., Aronson, J. A., Nys-trom, L. E., & Cohen, J. D. (2003). The neuralbasis of economic decision-making in the Ultima-tum Game. Science, 300, 1755–1758. http://dx.doi.org/10.1126/science.1082976

Schilbach, L., Eickhoff, S. B., Cieslik, E., Shah, N. J.,Fink, G. R., & Vogeley, K. (2011). Eyes on me:An fMRI study of the effects of social gaze onaction control. Social Cognitive and Affective Neu-roscience, 6, 393– 403. http://dx.doi.org/10.1093/scan/nsq067

Schleim, S., & Schirmann, S. (2011). Philosophicalimplications and multidisciplinary challenges ofmoral physiology. Trames, 15, 127–146. http://dx.doi.org/10.3176/tr.2011.2.02

Talairach, J., & Tournoux, P. (1988). Co-planar ste-reotactic atlas of the human brain: 3-dimensionalproportional system: An approach to cerebral im-aging. New York, NY: Thieme.

Tamm, L., Menon, V., & Reiss, A. L. (2002). Mat-uration of brain function associated with responseinhibition. Journal of the American Academy ofChild & Adolescent Psychiatry, 41, 1231–1238.

132 SMITH, ANAND, BENATTAYALLAH, AND HODGSON

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 18: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

http://dx.doi.org/10.1097/00004583-200210000-00013

Tversky, A., & Kahneman, D. (1981). The framing ofdecisions and the psychology of choice. Science,211, 453– 458. http://dx.doi.org/10.1126/science.7455683

Uddin, L. Q., Iacoboni, M., Lange, C., & Keenan,J. P. (2007). The self and social cognition: The roleof cortical midline structures and mirror neurons.Trends in Cognitive Sciences, 11, 153–157. http://dx.doi.org/10.1016/j.tics.2007.01.001

Woolrich, M. W., Behrens, T. E., Beckmann, C. F.,Jenkinson, M., & Smith, S. M. (2004). Multilevellinear modelling for FMRI group analysis usingBayesian inference. NeuroImage, 21, 1732–1747.http://dx.doi.org/10.1016/j.neuroimage.2003.12.023

Wright, P., He, G., Shapira, N. A., Goodman, W. K.,& Liu, Y. (2004). Disgust and the insula: FMRI

responses to pictures of mutilation and contamina-tion. Neuroreport, 15, 2347–2351. http://dx.doi.org/10.1097/00001756-200410250-00009

Young, L., & Koenigs, M. (2007). Investigating emo-tion in moral cognition: A review of evidence fromfunctional neuroimaging and neuropsychology.British Medical Bulletin, 84, 69–79. http://dx.doi.org/10.1093/bmb/ldm031

Zahn, R., Moll, J., Krueger, F., Huey, E. D., Gar-rido, G., & Grafman, J. (2007). Social conceptsare represented in the superior anterior temporalcortex. PNAS Proceedings of the National Acad-emy of Sciences of the United States of America,104, 6430 – 6435. http://dx.doi.org/10.1073/pnas.0607061104

Received March 8, 2013Revision received April 27, 2015

Accepted April 28, 2015 �

133FMRI AND HEALTHCARE DECISIONS

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z xppws S�1 6/7/15 23:19 Art: 2013-1039

APA NLM

Page 19: An fMRI Investigation of Moral Cognition in Healthcare Decision …eprints.lincoln.ac.uk/id/eprint/17743/1/Healthcare... · 2015. 6. 22. · Correspondence concerning this article

JOBNAME: AUTHOR QUERIES PAGE: 1 SESS: 1 OUTPUT: Sun Jun 7 23:19:27 2015/tapraid5/nec-npe/nec-npe/nec00215/nec0109d15z

AQau—Please confirm the given-names and surnames are identified properly by the colors.� Given-Name, � Surname

The colors are for proofing purposes only. The colors will not appear online or in print.

AQ1—Author: This article has been lightly edited for grammar, style, and usage. Please compareagainst your original document and make changes on these paged proofs. If no change isrequired in response to a question, please write “OK as set” in the margin.

AQ2—Author: Please be sure to provide the name of the department(s) with which you and yourcoauthors are affiliated at your respective institutes if you have not already done so. If you areaffiliated with a governmental department, business, hospital, clinic, VA center, or othernonuniversity-based institute, please provide the city and U.S. state (or the city, province, andcountry) in which the institute is based. Departments should be listed in the author footnoteonly, not the byline. If you or your coauthors have changed affiliations since the article waswritten, please include a separate note indicating the new department/affiliation: [author’sname] is now at [affiliation].

AQ3—Author: Please review the typeset table(s) carefully against your original table(s) to verifyaccuracy of editing and typesetting.

AQ5—Author: Please confirm figure(s) and legend(s) as set.

AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 1