J7ournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:709-7 16

Minor head injury: pathophysiological oriatrogenic sequelae?

Freda Newcombe, P Rabbitt, M Briggs

AbstractThis study addresses the possibility thatcognitive sequelae-albeit of a transientor minor character-can be associatedwith mild head injury. Twenty men (aged16-30 years of age), whose post-traumaticamnesia did not exceed eight hours, wereexamined within 48 hours of their acci-dent and again one month later. Thisunselected sample had no previous his-tory of head injury. A control group of 20men of similar socioeconomic back-ground, was selected from medical wards(where they had been admitted fororthopaedic treatment or a minor opera-tion). They were also retested one monthafter the first examination. Neuropsycho-logical tests were selected to measureabilities often compromised after signifi-cant head injury, namely memory andattention. The experimental componentconsisted of the fractionation of a com-plex skill (paced addition) to probe fordeficits at different stages of informationprocessing: perception and input intostorage; search for and retrieval of infor-mation from working memory; andpaced and unpaced addition. In general,no significant difference was foundbetween the experimental and controlgroups, with the possible exception of aninitial decrement on two working mem-ory tasks: probe digits and a keepingtrack task (where the subject has to keepin mind and update a number of vari-ables at the same time). The keepingtrack paradigm, ostensibly of ecologicalrelevance, may well be worth furtherexploration in memory research, and instudies of more severely head-injuredpatients. It is further suggested that theappropriate management and coun-selling of mildly head-injured patientsmay help to avert symptoms that are ofpsychological rather than pathophysio-logical origin.

(7 Neurol Neurosurg Psychiatry 1994;57:709-716)

The classical taxonomy' of blunt head injury,in terms of the severity of brain damage, isusually defined by the duration of post-trau-matic amnesia (PTA). In the absence of read-ily accessible and appropriate neuroimagingto reflect the massive axonal shearing associ-ated with such injury,2 PTA has beenconsidered the best available biological

marker of severity of brain damage.Meticulous necropsy data' on patients whodied from other causes, however, have indi-cated diffuse neuronal shearing after (very)minor head injury, and thus implicity widenedthe field of enquiry. Cases of minor headinjury, initially relegated to "the dumpingground of neurasthenia" (Symonds,4 p.607),were then entitled to neuropsychological eval-uation with the expectation that subtle formsof attentional or memory loss, or both, mightbe elicited.The seminal research of Gronwall and

coworkers'-7 indicated such attentionalimpairments, at least during the weeks imme-diately after injury. Neurophysiologicalchanges persisting up to six months after avery mild head injury have also beenreported8: large deviations from the normalevoked potential response were recorded inthese trauma patients as they worked at vigi-lance and discrimination tasks.

Nevertheless, there is (as yet) no clear con-sensus as to the nature and duration of thecognitive sequelae of minor head injury. Onelarge-scale and often quoted study (Rimel etal,9 p.221) suggested "some problems withattention, concentration, memory or judge-ment in most of the subset of 69 patients eval-uated neuropsychologically" about threemonths after injury. It appears, however, thatthe large sample (with which this subset ofpatients was comparable in terms of sex, edu-cation, and employment) included a largegroup of college students, a higher percentageof sporting injuries than is common in othersamples,'0 and a relatively high (35%) per-centage of patients with a blood alcohol levelin excess of the Virginian limit (0-10 g/l) at thetime of injury. Almost one-third gave a historyof previous head injury. Hence it is unwise toassume that the reported sequelae are charac-teristic of the effects of a single head injury,uncomplicated by other factors, in othergroups of varied demography.

Significant cognitive changes, however,have been reported"-"3-at least on short-termfollow up-after relatively minor head injury(variously defined as a PTA of less than oneday, absence of coma, or brief admission tohospital). Such changes are difficult to evaluatein the absence of appropriate control proce-dures.

In contrast, several studies of minor headinjury"4-17 have provided essentially negativeevidence, at least with regard to permanentsymptoms. Gronwall and Wrightson,6 forexample, reported recovery of performance on

Department ofNeurosurgery, TheRadcliffe Infirmary,Oxford, UKF NewcombeM BriggsAge and CognitivePerformanceResearch Centre,University ofManchester,Manchester, UKP RabbittCorrespondence to:Dr F Newcombe,Departnent ofNeurosurgery, RadcliffeInfirmary, Oxford OX26HE, UK.Received 15 April 1993and in revised form16 August 1993.Accepted 24 August 1993

709 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Newcombe, Rabbitt, Bniggs

a paced addition task (PASAT) 54 days afterinjury in 75 (94%) of 80 patients who hadbeen briefly concussed. In a study in Belfast,'5strengthened by the inclusion of a controlgroup, patients reached at least the controllevel of performance on a four-choice reactiontime task six months after injury, despite theirsignificantly poorer performance after shorterintervals of 24 hours and six weeks. The PTAwas less than 24 hours and, for half of thegroup, less than 15 minutes; "those sufferingfrom alcohol intoxication or any other compli-cated illnesses were excluded".'5

Likewise, no significant impairment wasfound on a variety of cognitive tests (digit can-cellation, digit span, verbal memory, wordrecognition, Buschke's selective remindingparadigm for word recall, card sorting, andRaven's progressive matrices) in a group of 50Italian patients tested one month afterinjury.'6 There were clearly defined criteria foradmission to this study: loss of consciousnessfor less than 20 minutes and an initial scoreon the Glasgow coma scale'8 of 13-15.Patients selected their own controls fromamong family, friends, and schoolmates("case control pairing").

Similarly, a carefully designed three-centrestudy in the USA suggested that "a single,uncomplicated minor head injury producesno permanent, disabling neurobehaviouralimpairment in the great majority of patientsfree of pre-existing neuropsychiatric disorderand substance abuse" (Levin et al,'9 p.234).Of the 155 patients initially studied, 57 (37%)returned for a one-month follow up and metthe strict criteria for inclusion in the study.These criteria included a Glasgow coma scalescore of at least 13, and absence of focal neu-rological deficit, intracranial mass lesion(haematoma), intracranial surgical procedures(for example, repair of depressed skull frac-ture), or complications such as meningitis.Other criteria included admission to hospitalfor 24 hours or less, no general anaesthesia(for example, for an operation related toextracranial injuries), and no history ofantecedent neurological disorder, admissionto hospital for previous head trauma, sus-tained alcohol or drug abuse or admission tohospital for psychiatric disorder. The baselineassessment was carried out within seven daysof injury when the initial PTA had resolved,and the follow-up examination no later thanfive weeks after injury. The tests were selectedto measure attention, memory, and speed ofinformation processing. The three-centrestudy strongly supported the view that theserigorously selected patients with minorclosed-head injury did not differ from well-matched controls on tests carefully chosen toprobe those skills that are characteristicallyimpaired in more severe closed-head injury. Italso provided striking evidence of epidemio-logical influences: considerable variationbetween centres with respect to the aetiologyof injury and a remarkable difference in thelevel of performance between centres. On ini-tial testing, the control group was superior tothe head-injured group on all five tasks of

which the verbal memory test was the mostsensitive. This group difference disappeared,however, one month after injury when therewas a notable improvement in verbal memory.Post-concussional symptoms-headache,dizziness, and decreased energy-were ini-tially reported by a substantial number of thetotal patient sample but had decreased at theone-month follow up to 56, 35, and 47%,respectively. These workers note that"patients treated in emergency rooms or hos-pitalized for minor head injury are typicallygiven minimal or no instructions concerningthe risk for development of neurobehaviouralsequelae and temporal course, nor are theseresidual effects adequately evaluated on anysystematic basis" and they suggest that "sec-ondary emotional distress leading to aggrava-tion of post-concussional symptoms andexcessive time away from work after minorhead injury could be mitigated by early clini-cal intervention..." (Levin et al,'9 pp.241-3).

Given the controversial nature of thereported research on the sequelae of minorhead injury, we decided to test the possibilityof subtle deficits in memory, attention, andspeed of processing in a group of young menwith a PTA of less than one day. The samplewas particularly suitable for this exercise. TheOxford accident service admits over 1000patients with trauma annually; these are directreferrals from a wide catchment area andhence compose an unbiased sample of traumapatients in three counties. A pilot study (citedin Thomas20) established the main demo-graphic characteristics of the 340 menbetween the ages of 16 and 55 admitted duringa six month period. Road traffic accidentsconstituted the major cause of trauma (46%),followed by falls (16%), and assaults (13%).Over 80% were estimated to have had PTA ofless than one day and over 90% were dis-charged within 24-48 hours. Against thebackground of the national statistics avail-able,'0 the population provides a reasonablecross section of trauma patients in the UK,without the notorious alcohol intake of a dis-proportionate number of assault cases to befound in a few metropolitan areas. A subset ofsubsequent admissions from our regionalpopulation was therefore considered emi-nently suitable for a study of the cognitivesequelae of a first, minor head injury.

In applied research of this nature there arealternative but complementary researchstrategies. The traditional psychometricapproach analyses performance on measuresof those cognitive abilities that are consis-tently reported to be impaired after headinjury: namely, memory and attention.21-26Alternatively, a complex task (reportedly sen-sitive to the effects of closed-head injury) canbe fractionated in terms of an informationprocessing model. Predictions can then bemade about performance from this model andtested with appropriately designed experimen-tal tasks. In the absence of a solid foundationof prior knowledge about the sequelae ofminor head injury, we adopted bothapproaches.

710 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Minor head injury: pathophysiological or iatrogenic sequelae?

Subjects and methodsSUBJECTSTwenty men between the ages of 16 and 30years who lived within easy reach of the hospi-tal and who were admitted to the accidentservice after minor head injury betweenFebruary and August 1976 were examined.The injuries were due to road traffic acci-dents; no skull fractures were present. In nopatient was there a previous history of headinjury and none had physical injuries suffi-ciently severe to prevent them getting out ofbed within the first 48 hours of injury.Estimates of the length ofPTA were based onambulance and clinical notes, interview, andthe continuous screening of memory and ori-entation by the neuropsychologist. The num-ber of patients in the different categories ofPTA were as follows: PTA of a few minutes,seven (35%); less than half an hour, seven(35%); less than one hour, three (15%); andone to eight hours, three (15%). Patients weretested within 48 hours of the accident andagain one month later.Twenty male control patients were selected

from the medical wards: nine orthopaedicpatients and 11 patients admitted for a minoroperation. They were tested within one weekof admission and then again at follow up onemonth after admission.The two groups did not differ in age or per-

formance on standard verbal and non-verbaltests of intelligence: head-injured group mean(SD) age 21A4 (3-9), matrices score 45-7(11.1), and Mill Hill IQ equivalent 99 3(14-7); control subjects mean (SD) age 21-2years (3-5), matrices score 49-6 (6-4), andMill Hill IQ equivalent 105-2 (12-8).

TEST PROCEDUREThe two groups were each tested twice: theexperimental group within 24 hours of admis-sion and one month later; the control groupon two occasions, initially during hospitaladmission and, at follow up, one month later.Owing to medical and administrative exi-

gences, it was, in a few instances, not possibleto administer all the tests. In the two groups,however, all but two subjects completed allthe tests, with the sole exception of the con-trol test of arithmetical ability which was notadministered to three members of the head-injured group.

TEST SELECTIONMemory and information processingTests of memory and information processingoften produce significant differences betweenpatients with severe closed-head injury andnormal controls. They were thereforeincluded on the premise that smaller but con-sistent differences might be detected after lesssevere injuries. In addition, one verbal mem-ory task-story recall-is known to correlatewell with relatives' ratings of the daily lifememory25 of the head-injured patient. Thefollowing tests were administered.

Story recallBower's27 hierarchical story-frame paradigm

was used to test the possibility that the head-injured patients would either be unable toabstract the hierarchy or forget more details atlower levels of the hierarchy. This procedureawards higher scores for recalling generalabstract features of the story and lower scoresfor details-unlike conventional story recalltests in standard test batteries. The story usedfor the first test session was a shortened ver-sion of Circle Island with 194 words and 32recall items. The (Torrida) story for the sec-ond test session was devised to approximatethe Circle Island format, with 199 words and33 recall items. At each session the experi-menter read the story aloud and the subject'sverbatim response was recorded on tape.Delayed recall, without forwarning, wasrecorded one hour later. The score was thenumber of story items recalled correctly.

Word andface recognitionVerbal and non-verbal continuous recognitiontasks, based on the Shepard andTeghtsoonian paradigm,28 were designed andincluded to check the rate of loss of items inimmediate memory as a function of thepassage of time and intervening interference.Such tasks might therefore detect a factorimplicated in the previously reported difficul-ties with paced addition tasks.S7 Subjectswere shown a set of stimuli and told that eachstimulus occurred twice in the test pack withvarying numbers of other stimuli between thetarget items. They had to go through the packand respond "old" if the stimuli had beenseen before in the pack and "new" if it hadnot occurred previously. There were fiveintervals, with stimuli recurring after one, five,10, 15, or 20 cards.

In the verbal version of the test, the stimuliconsisted of six-letter words typed on individ-ual cards. The 50 target stimuli consisted of25 high frequency (Thorndike-Lorge AA) and25 low frequency (Thorndike-Lorge 5/mil-lion) words. Five words of each frequencyoccurred at each interval. Three words whichdid not recur were included at the end of thepack to keep the probability of new or oldstimuli at 50%. There were thus 53 newwords which could have been falsely recog-nised. A brief practice task was given initiallyto ensure that all the subjects understood thetest.

In the non-verbal version of the task, thestimuli consisted of 40 pictures of the faces ofyoung ex-servicemen, with neutral expres-sions, and without paraphernalia or idiosyn-cratic hair cut. There were eight target itemsat each of the five intervals and 44 opportuni-ties for false positive errors.

Prospective memoryAt the end of the first test session each patientwas given a form to fill in when he returned towork and a stamped addressed envelope inwhich to return it to the experimenter. Thenumber of patients who returned the slip andthe time elapsing between return to work andreturn of the slip (the date of the postmark)were taken into account.

71 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Newcombe, Rabbitt, Bniggs

Card sortingSorting tasks were included to provide a com-parative measure of the rate of informationprocessing between patients and controls.The subject was given a pack of cards facedown and was asked to sort them in labelledred and black piles. For the control measure-ment of movement time, subjects were given apack of alternating red and black cards to besorted into the two piles. In the two-choicecondition, they were given a shuffled pack ofcards, again to be sorted into red and blackpiles. This test was repeated three times withthe left/right spatial position of the red andblack pile alternating on each occasion. Timeswere averaged. In the four-choice condition,the shuffled pack had again to be sorted intotwo piles but each pile contained two suits -that is, hearts and spades on one side and dia-monds and clubs on the other. This test wasrepeated four times with each of the two com-binations of suits occurring twice, each timein an alternate spatial position. Time scoreswere averaged. At the beginning of each ses-sion, packs were checked so that there werenever more than three of the same colour andnot more than one of the same suit in succes-sion. Between each test, packs were shuffledat least four times. The two-choice and four-choice conditions were repeated once at theend of the testing programme as a check forpossible fatigue effects.

PACED ADDMTION TASK AND ITS COMPONENTSThe paced addition task (PASAT)7 29 30 wasthe key experimental paradigm, selected onthe basis of Gronwell and Sampson's5 innova-tive studies. In this task, subjects hear astream of single digits. Their task is continu-ously to add the last digit spoken to the oneimmediately preceding it. The speed ofpresentation can be varied. Gronwell andSampson5 ascribed the reduced cognitiveefficiency of patients during the first monthafter mild and moderate concussion (reflectedon PASAT, choice reaction time, and speechprocessing tasks) to "reduced information-processing capacity" (p. 84). They eliminatedselective attention, response production, andmovement time as critical parameters.

In analysing the PASAT task we envisagedthree main stages of information processingand the experimental techniques that mightbe required to test these components.

Perception and input into storageThis component was thoroughly explored andeliminated by Gronwall and Sampson5 as acritical parameter.

Search for and retrieval of information fromworking memoryIn the PASAT task the correct (penultimate)digit must be selected from several items inworking memory, and then added to the lastdigit presented. As a reasonable simulation ofthis component of PASAT, we chose theprobe digit technique.3" Here, subjects have toidentify items in working memory by theirnature (item search).

In addition, the efficiency with which sub-jects could keep track of information held inworking memory was also tested by theYntema and Mueser paradigm.'2 The lattertask checks the rate of loss of items in immedi-ate memory as a function of the passage oftime and intervening interference. It thusinvestigates a component of PASAT to findout whether the rate of loss of items in memoryis also implicated in PASAT failures.

Paced and unpaced additionStraightforward control tasks measured per-formance on the arithmetical component ofPASAT when there was no extra demand onresources by a working memory load. Thespecific tests selected are described in the fol-lowing.

Modified PASAT-The present versionconsisted of 51 digits (50 pairs of single digitsto be added together) relayed on a portableSony tape recorder at two different speeds:'slow' (one digit every three seconds) and'fast' (one digit every 1.5 seconds). Omissionsand four types of error were recorded toattempt to check the relative vulnerability tohead injury of the various cognitive processesinvolved in this complex task. Omissions andlate errors might indicate breakdowns due tothe optimum speed of performance beingexceeded. "Retrieval errors" and "additionerrors" might indicate inefficient addressingor maintenance of information in workingmemory. "Miscellaneous errors" representeda catch-all category of failures unattributableto any particular process. Two short control/practice tasks preceded the administration ofPASAT: the successive repetition of 10 digitsspoken by the same experimenter and relayedon the same apparatus and the addition ofdigit pairs, presented at the "slow" rate (onedigit every three seconds).

Arithmetical ability test-To check parity onthose components of PASAT which involvedpaced arithmetical ability subjects were askedto add pairs of single digits (e.g. 8 + 3 = ?)played on the same tape recorder as the con-ventional PASAT at the same two speeds.Twenty-five additions were given at the "slow"speed and 50 at the "fast" speed. Incorrectresponses were scored as "late responses","omissions", or "additional errors".

Probe digits test-These tasks checked theability to retrieve items and item order fromworking memory. Subjects heard a series ofseven random digits relayed on a taperecorder, followed, after a three second gap,by a single digit. There were two conditions.In the item condition, subjects had to saywhether the single (probe) digit had occurredin the immediately preceding set of seven dig-its. Twenty four positive and 24 negativeprobes were given in pseudo-random order.Digits in each of the six positions (exceptingthe first) were probed twice. In the order con-dition, the probe digit had invariably occurredin the preceding series of seven digits. Thesubject had to name the digit that had imme-diately preceded the probe. Each of the sixpositions in the series (excluding the first) was

712 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Minor head injury: pathophysiological or iatrogenic sequelae?

tested four times. The conventional scoringsystem was used.

Keeping track task-This task was based onan interesting paradigm developed by Yntemaand Mueser32 to assess the subject's ability tokeep track of several variables, simultane-ously. Ostensibly, it might tap a proverbialdifficulty of the head-injured patient inattending to several concurrent strands ofconversation. One of the problems in PASATis that subjects have to select the correct oneof several possible digits, correctly referencingitems in working memory. The keeping tracktask was therefore an extended analogue ofthis aspect of the PASAT task. Subjects haveto address items not merely in terms of tem-poral sequence but also in terms of category(pile).

Patients were asked to sort, face down, apack of 56 cards into four piles according tocategory: digits (1-10), letters (A-K, exclud-ing I), colours (colour names written in theappropriate colour in upper case letters), andshapes (heart, diamond, club, spade, circle,square, rectangle, triangle, star and cross, out-lined in black ink). There were 10 messagecards in each category and the message cardswere interspersed with question cards, dis-playing a Q and the name of the category.Subjects then had to recall as many itemsfrom that category as they had previouslysorted. The number of cards interveningbetween the presentation of an item and itssubsequent recall varied from none to seven;and there was one item at each interval fromboth the "verbal" (number and letters) and"non-verbal" (colours and shapes) category.The score was taken as the number of itemscorrectly recalled (maximum 76, 19 fromeach category). Errors included extraneousitems (for example, letters beyond K in thealphabetic sequence) and visual confusions.

ResultsINITIAL TESTINGSignificant group differences were observedonly on working memory tasks-keeping trackand item digit probe. Other group compar-isons were negative.

In the interests of brevity, analyses of vari-ance for specific tasks are summarised in thefollowing.

Memory and information-processing paradigmsStory recall-There was a significant effect

of delay (F = 9.93, p < 0-005) for the twogroups (see tables 1 and 2). To investigateBower's32 centrality hypothesis, the first andsecond higher-order levels in the Circle Islandstory hierarchies were combined and com-pared with the third and fourth levels. TheANOVA took delays and levels as repeatedmeasures to be computed for each story sepa-rately. The effect of levels was significant(F = 37 43, p < 0 005) as was the effect ofdelays (F = 7-25, p < 0.01), but there was nodifference between the two groups. A similarpattern emerged from an analysis of theTorrida story.

Recognition-The data for the verbal versionwere subjected to a three-way analysis ofvariance consisting of groups, frequencies,and intervals between successive presenta-tions of the stimulus. The main effects ofinterval (F = 4-48, p < 0 005) and frequency(F = 4.77, p < 0-05) were significant, butthere was no difference between the groups.In addition, there was no difference betweenthe groups in terms of error pattern (false neg-ative and false positive errors). For the twogroups, performance was worse on high fre-quency words and with longer intervalsbetween successive presentations of the samestimulus.

In the non-verbal (faces) version only theeffect of interval was significant (F = 4 56,p < 0005) for the two groups.

Prospective memory-The difference be-tween the percentage of head-injured (60%)and control (65%) patients returning theirforms did not differ significantly (x2 = 0-008,df = 1). The time elapsing between return towork and return of the form did not differbetween the two groups (mean 3-5 days),excluding one control patient who broughthis form back 10 days before returning towork.

Card sorting-The effect of choice washighly significant (F = 465-21, p < 0 0005):sorting was slower if there were more alterna-tive categories. There was no significant dif-ference between the groups, though thehead-injured patients were on average 5'7 sec-onds slower than the control patients and alsosignificantly more variable (tests for homo-geneity of variance: movement time F = 2-61,p < 0-05; two choice, Fma, = 3-27, p < 0-025;four choice, Fm. = 3 40, p < 0-05).

Paced addition and its componentsModified PASAT-The main effect of error

type was significant (F = 3-36, p < 0 025) andthere was a significant interaction betweenerror type and speed (F = 6-73, p < 0-0005).The two groups made more late responses onthe fast speed than on the slow speed andmore of the other error types on the slowerspeed. This interaction cancelled out the maineffect of speed.

Paced addition-There was a highly signi-ficant effect of speed as a repeated measure(F= 178-19, p < 0-001): the two groups per-formed less well at the faster presentationrate. There was, however, no significant dif-ference between groups nor interactionbetween groups and presentation rate.

Probe digits-There was a highly significanteffect of position (F = 6-70, p < 0-0005) inthe item condition: performance was best onthe last three positions, with fewer items inter-vening between presentation and recall; it wasalso relatively good on the first item of theseries. The head-injured patients recalledsignificantly fewer items than the controls(F=810, p < 0-005). Tables 3 and 4 givethese results.

Effect of position in the order conditionwas highly significant (F = 5-76, p < 0-0005):performance was best on the first and last

713 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Newcombe, Rabbitt, BriggsTable 1 Keeping track task: mean (SD) percentagerecall

Head-injured ControlTask category patients subjects

Numbers 36-3 (25 4) 48-4 (18-5)Letters 25-0 (13-6) 38-7 (20 6)Shapes 66-3 (17-4) 69-2 (11-6)Colours 56-3 (21-2) 65-8 (20 8)Total 46-0 (14-1) 55-5 (14-1)Time (s) 371-5 (116-8) 331-1 (82 2)

Table 2 Keeping track task: analyses of vaniance

Source

df MS F

Groups 1 131-41 997Categories 3 396-31 30-06Groups x categories 3 8-17 0 62Between subjects within groups 152 13 18

items of the series. There was, however, no

group difference.Keeping track-The head-injured patients

recalled significantly fewer items than thecontrols and the effect of category was alsosignificant (F = 30-06, p < 0 0005) but therewas no interaction effect (tables 1 and 2). Thehead-injured patients differed significantlyfrom the controls on letter (t = 2-48, df = 33,p < 0-01) and number recall (t = 1-72,df = 33, p < 0-05) as well as on total recall(t 2-4, df = 33, p <0Q025). The twogroups, however, reported that the letters andnumbers were more difficult to recall thancolours and shapes, an effect probably attrib-utable to set size and imageability. The twogroups did not differ significantly in sortingspeed (t = 1-18, df = 33, p < 0 25).

Table 3 Item probe task: mean (SD) percentage correct

Head-injuredItem conditions patients Control subjectsPosition

2 89 0 (21 4) 89-5 (26 8)3 66-5 (38 3) 84-0 (40-1)4 72-0 (31-1) 94-5 (15-8)5 94-0 (16-2) 100-0 (0 0)6 91-5 (19-2) 94 5 (15-0)7 100-0 (0 0) 100-0 (0 0)

Order2 63-8 (31-2) 75 0 (27 6)3 45-8 (31-2) 46-0 (39*3)4 55-5 (27 8) 54-0 (33 6)5 30 5 (20 3) 51-25 (34 8)6 47-3 (28 3) 55-25 (284)7 72-3 (29 6) 68 5 (24 8)

Total 52 8 (21-3) 58-3 (24 8)

Table 4 Item probe task: analyses of variance

Source

df MS F

Item conditionGroups 1 1-50 8-10Positions 5 1-23 6-70Groups x positions 5 0 33 1-81Between subjects within groups 210 0-18

Order conditionGroups 1 2-97 1-97Positions 5 8-71 5-76Groups x positions 5 1-29 0-85Between subjects within groups 210 1-51

Homogeneity ofsampleOn the assumption that the head-injuredgroup might be a heterogenous sample,including a subset of patients who were signif-icantly impaired by the accident, the datawere analysed in two ways: a search for con-sistently poor performance in a subset ofpatients and a correlation between perfor-mance and grade of PTA.

There was no evidence of a subset ofpatients showing impairment on all tasks. Inrelation to PTA, only broad categories couldbe used because of the restricted range andinevitable imprecision of measurement.

Three categories of PTA were taken intoconsideration: less than five minutes (sevenpatients), between five and 30 minutes (sevenpatients),- and more than 30 minutes (sixpatients). No significant relation was foundbetween task performance and duration ofPTA within these narrow limits. The sub-group with the longer time of PTA obtainedmarginally higher scores on the two intelli-gence tests (though this effect did not reachsignificance on an analysis of variance).

There was, however, a consistent trendtoward increased variability in the head-injured group which reached significance incard sorting.

FOLLOW UPSmaller groups of head-injured (14 patients)and control (14 patients) subjects returned forfollow up. There was again a significant differ-ence between the two groups on the itemcondition of the probe digit task (F = 10-95,p < 0.005). Otherwise, there were few signifi-cant changes in the pattern of performance.These consisted of: PASAT (the two groupsscored significantly higher on retest, at slow(F = 16-88, p < 00005), medium (F = 8-03,p < 0.025), and fast (F = 2083, p < 0 0005)speeds); story recall (the effect of delay was nolonger significant-the two groups remem-bered as much of the story an hour later asthey had on immediate recall); and keepingtrack (there was now no difference in meanscore between the two groups).

DiscussionIn an unselected and hence representativesample of patients with minor head injury inthe Oxford region, there was no evidence of asignificant and overall decrement in perfor-mance on cognitive tests within 48 hours ofinjury compared with that of control subjects.Had there been any substantial cognitiveimpairment, we would have expected to haveelicited it with the tasks selected and within ashort time interval after injury. The absence ofan overall group difference is perhaps surpris-ing in view of "the clear evidence of cognitivedeficits in the first few days after mild headinjury", listed by Binder"3 (p. 327). Our osten-sibly negative findings are, however, consis-tent with the European studies previouslycited, including those of Lidvall et al17 whofound no cognitive impairment in mildlyhead-injured patients tested two, six, 14, 30,

714 on A

pril 4, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. Dow

nloaded from

Minor head injury: pathophysiological or iatrogenic sequelae?

and 90 days after injury and also with one ofthe most thorough and rigorous studies inNorth America by Levin and colleagues.'9Demographic and cultural factors, as well asdifferences in medicolegal practice, may per-haps play an important part in explaining suchreported discrepancies. There are, of course,other factors to be taken into account includingthe age of the patients: older patients arethought to be "at greater risk for prolongeddisability"'13 (p.328). In addition, there arealways problems in the selection of appropri-ate control subjects. In this study the fact thatsome of the controls had received an anaes-thetic (albeit mild and at least 24 hours beforetesting) might, arguably, have reduced possi-ble experimental and control group differ-ences. The control group did not performsignificantly better, however, when retestedone month after treatment.Our inability to replicate Gronwall and

Sampson's5 findings could be attributed to anumber of factors. Major factors were probablythe shorter trials and slower pacing rates used.Gronwall's (personal communication) reanaly-sis of her original data, using only the first 50additions, showed that only two of the speedconditions (digits presented at 0-8 and 2-4second intervals) produced significant differ-ences between the experimental and controlgroups. There were also procedural andmethodological differences between the twostudies. Half the subjects in the earlier studyreceived the fast version of PASAT first.Then, in this study, the 3 0 s trial is slowerthan any of those used in Gronwall andSampson's original studies.5 Another factor tobe taken into account is the length of theexamination. In this experiment, the subjectsworked continuously for about two hours. It isconceivable that a longer period of testingmight have elicited fatigue and reduced levelsof competence in the experimental group, butGronwall reported that when a 4'0 secondinterval was introduced between digits, head-injured patients were better than controls,though only a small group (six patients) wasused and the results were not significant. Wenote, nevertheless, that Oddy33 did not find adifference between groups of controls andyoung male head-injured patients (selected onthe basis of a PTA of more than two days)tested six months after injury. It has also beensuggested that the PASAT may be an unreli-able task, showing considerable intra-subjectdifferences.Two aspects of our data, however, suggest

caution in disclaiming any significant, even iftransitory, effects of minor head injury. First,there was an early decrement on tasks mea-suring working memory (digit probe andkeeping track), and the significant differenceon digit probe was found again one monthafter injury (albeit in small groups due to the30% dropout rate in experimental subjectsand controls). It could be argued that thesefew significant differences emerged by chance.In rebuttal, they can be considered to have acertain functional consistency. Thus measure-ment of three distinct processes-information

processing rate, intermediate term memory,and working memory-showed no decrementexcept for the mild decrease in working mem-ory, at least in the initial examination. Hencewe suggest that although it has been very hardto produce convincing evidence of impair-ment in a number of the well-controlledrecent studies, it is possible that selectivestudies of working memory would tap signifi-cant effects.The possibility of group differences is sup-

ported by data from a study parallel with ourown by Saan (unpublished data) in whichanother unselected group of young men witha first, single head injury was compared withother well-matched control subjects.Embedded in a cluster of negative findings(from tasks including writing and tappingspeeds, letter and design fluency, letter anddot cancellation, and sentence comprehen-sion) was a significant group difference onanother version of the keeping track task.What might have been considered a statisticalartefact tends to reinforce the notion of a(probably transient) working memory deficit.We suggest that the keeping track task pre-sents an interesting paradigm that may welltap a typical patient disorder: keeping in mindand updating a number of variables at thesame time-perhaps not dissimilar to theproverbial "cocktail party problem" of theaphasic patient.

Secondly, the consistent increase in thevariance of the head-injured patients' perfor-mance (see tables 3 and 4, a modified sign teston the variance was significant at the level ofp < 0.01) hints at the possibility of individualoutliers whose performance is impaired (seealso Leininger et al34) . Again, this findingraises the possibility of subtle, selectiveeffects; variability of performance may be amore sensitive index than rate alone.

Future studies of minor and moderate headinjury call for more sophisticated test proce-dures35 36 and more heuristic models ofmemoryand attentional processes37 38 than have beenavailable or used in the past. Given that back-ground, and a rigorous definition of patientand control samples, uncontaminated by sub-stance abuse and medicolegal manipulation,we may be better equipped to study the "pri-mary" effects of closed (blunt) head injury.Recent advances in neuroimaging-specifi-cally, the superiority of MRI over CT inreflecting CNS damage after head injury343-will help increasingly to define the nature andextent of CNS damage in relation to behav-ioural (cognitive and affective) changes.Sophisticated neuroimaging may also serve as acomplementary measure of severity of damageto that of PTA which, as Symonds4 (p.606)pointed out over 40 years ago, can be mislead-ing, and is often not systematically measured(but see references 44-46).The delicate issue of primary symptoms

and secondary psychosomatic, emotional, ormotivational factors will not be broached indetail (for judicious reviews see references 43and 44). We will only recapitulate that manypatients with closed-head injury are given

715

on April 4, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. D

ownloaded from

Newcombe, Rabbitt, Briggs

little in the way of positive counselling,47 48despite the reported efficacy of early reassur-

ance and proactive treatment.49 A carefulstudy by Minderhoud et al,50 which shouldcertainly be replicated, contrasted the recov-

ery from minor head injury of unselectedgroups of patients: one group was not treatedand was not usually given an outpatient fol-low-up appointment. The experimental group

was "treated with strict instructions regardingthe period they should remain in bed .... andgiven as much information and encourage-

ment as was possible in the given circum-stances" (p.128). Minderhoud et al reportedthat post-concussional sequelae were

markedly reduced by treatment whichincluded information, explanation, andencouragement. They hypothesised that"post-concussional sequelae start off on an

organic basis (PTA) and that persistentsequelae after minor head injuries are alsocaused by psychogenic, and especially byiatrogenic factors" (p.127). In the manage-

ment of minor head injury, the clinician haspresumably to steer a judicious course

between positive counselling for the long term

and the detection of selective (usually tran-sient) cognitive impairments in individualpatients.

The authors are indebted to Ms Caroline MThomas for hermeticulous collection and analysis of data. The work was sup-

ported by MRC Grant No G/9731144 to Dr FredaNewcombe.

Russell WR.The traumatic amnesias. London: OxfordUniversity Press, 1971.

2 Teasdale G, Mendelow D. Pathophysiology of headinjuries. In: Brooks N, ed. Closed head injury: psychological,social and family consequences. Oxford: Oxford UniversityPress, 1984:4-36.

3 Oppenheimer DR. Microscopic lesions in the brain follow-ing head injury. J Neurol Neurosurg Psychiatry 1968;31:299-306.

4 Symonds CP. Discussion on differential diagnosis andtreatment of post-contusional states.Proc R Soc Med1942;35:601-14.

5 Gronwall DMA, Sampson H.The psychological effects ofconcussion. Auckland: Auckland University Press, 1974.

6 Gronwall D, Wrightson P. Delayed recovery of intellectualfunction after minor head injury. Lancet 1974;ii:605-9.

7 Gronwall D. Paced auditory serial-addition task: a mea-

sure of recovery from concussion. Percept Mot Skils1977;44:367-73.

8 Curry H, Cummins BH. Electrophysiological changesafter head injury. Presented at the Seventh InternationalCongress of Neurological Surgery; 1981 Jul 12-18;Munich.

9 Rimel RW, Giordani B, Barth JT, BollTJ, Jane JA.Disability caused by minor head injury. Neurosurgery1981;9:221-8.

10 Field JH. Epidemiology ofhead injuries in England and Wales.London: HMSO, 1976.

11 Van Zomeren AH, Deelm an BG. Long-term recovery ofvisual reaction time after closed head injury.Y NeurolNeurosurg Psychiatry 1978;41:452-7.

12 BarthJT, Macciocchi SN, Giordani B, Rimel R, Jane JA,BollTJ. Neuropsychological sequelae of minor headinjury. Neurosurgery 1983;13:529-33.

13 Binder LM. Persisting symptoms after mild head injury: a

review of the postconcussive syndrome. J Clin ExpNeuropsychol 1986;8:323-46.

14 McMillan TM, Glucksman EE. The neuropsychology ofmoderate head injury.Jf Neurol Neurosurg Psychiatry1987;50:393-7.

15 MacFlynn G, Montgomery E, Fenton GW, Rutherford W.Measurement of reaction time following minor headinjury. J Neurol Neurosurg Psychiatry 1984;47: 1326-31.

16 Gentilini M, Nichelli P, Schoenhuber R, et al.Neuropsychological evaluation of mild head injury. JNeurol Neurosurg Psychiatry 1985;48:137-40.

17 Lidvall HF, Lindroth B, Norlin B. Causes of the postcon-

cussional syndrome. Acta Neurol Scand Suppl1974;50:Suppl. 56, 1-144.

18 Teasdale G, Jennett B. Assessment of coma and impairedconsciousness. A practical scale. Lancet 1974;H:81-4.

19 Levin HS, Mattis S, Ruff RM, et al. Neurobehaviouraloutcome following minor head injury: a three-centrestudy. J Neurosurg 1987;66:234-43.

20 Thomas CM. Deficits of memory and attention followingclosed head injury. [unpublished MSc thesis]. Oxford:Oxford Univ, 1977.

21 Ruesch J. Intellectual impairment in head injuries. Am JfPsychiatry 1944;100:480-96.

22 Fodor IE. Impairment of memory function after acutehead injury. Jf Neurol Neurosurg Psychiatry 1972;35:818-24.

23 Brooks DN. Long and short term memory in head injuredpatients. Cortex 1975;1:329-40.

24 Brooks DN, ed. Closed head injury: psychological, social andfamily consequences. Oxford: Oxford University Press,1984.

25 Baddeley A, Sunderland A, HarrisJ. How well do labora-tory-based psychological tests predict patients' perfor-mance outside the laboratory? In: CorklnS, Davies KL,Growden JH, Usdin E, Wurtman RJH, eds. Alzheimer'sdisease: a report of progress. Vol 19. New York: RavenPress, 1982;141-8.

26 Deelman BG, Saan RJ, van Zomeren AH. Traumaticbrain injury. Clinical, social and rehabilitational aspects.Amsterdam: Swets & Zeitlinger, 1990.

27 Bower G. Experiments on story understanding and recall.QJExp Psychol 1976;28:511-34.

28 Shepard RN, Teghtsoonian M. Retention of informationunder conditions approaching a steady state. J ExpPsychol 1961;62:302-9.

29 Sampson H. Pacing and performance on a serial additiontest. Can JPsychol 1956;1O:219-25.

30 Sampson H. Stimulus duration and paced performance.Can JPsychol 1958;12:7-12.

31 Waugh NC, Norman DA. Primary memory. Psychol Rev1965;72:89-104.

32 Yntema DB, Mueser GE. Remembering the present stateof a number of variables.Y Exp Psychol 1960;60: 18-22.

33 Oddy MJ. Psychological and social aspects of recoveryfrom closed head injury [unpublished PhD thesis].London: Univ of London.

34 Leininger BE, Gramling SE, Farrell AD, Kreutzer JS, PeckEA. Neuropsychological deficits in symptomatic minorhead injury patients after concussion and mild concus-sion. J Neurol NeurosurgPsychiatry 1990;53:293-6.

35 Williams DH, Levin HS, Eisenberg HM.Mild head injuryclassification. Neurosurgery 1990;27:422-8.

36 Hugenholtz H, Stuss DT, StethemIL, Richard MT. Howlong does it take to recover from a mild concussion?Neurosurgery 1988;22:853-8.

37 Van Zomeren AH, Brouwer WH. Head injury and con-cepts of attention. In: Levin HS, Grafman J, EisenbergHM, eds. Neurobehavioural recoveryfrom head injury.New York: Oxford, 398-415.

38 Parasuraman R, Mutter SA, Molloy R. Sustained attentionfollowing mild closed head injury. Y Clin ExpNeuropsychol 1991;13:789-81 1.

39 Han JS, Kaufman B, AlfidiRJ, et al. Head trauma evalu-ated by magnetic resonance and computed tomography: acomparison. Radiology 1984;150:71-7.

40 Jenkins A, Teasdale G, Hadley MDM, Macpherson P,Rowan JO. Brain lesions detected by magnetic reso-nance imaging in mild and severe head injuries. Lancet1986;ii:445-6.

41 Levin HS, Handel SF, Goldman AM, Eisenberg HM.Magnetic resonanceimaging after 'diffuse' nonmissilehead injury. Arch Neurol 1985;42:963-8.

42 Snow RB, Zimmerman RD, Gandy SE, Deck MDF.Comparison of magnetic resonanceimaging and com-puter tomography in the evaluation of head injury.Neurosurgery 1986;18:45-52.

43 WilsonJTI, Wiedman KD, Hadley DM, Condon B,TeasdaleG, Brooks DN. Early and late magnetic reso-nance imaging and neuropsychological outcome afterhead injury. J Neurol Neurosurg Psychiatry 1988;51:391-6.

44 Artiola I, Fortuny L, Briggs M, Newcombe F, Ratcliff G,Thomas C. Measuring the duration of post-traumaticamnesia.J7 Neurol Neurosurg Psychiatry 1980;43:377-9.

45 Levin HS, Benton AL, Grossman RG. Neurobehavioralconsequences of head injury. New York: Oxford UniversityPress, 1982.

46 Shores EA, Marosszeky JE, Sandanam J, Batchelor J.Preliminary validation of a clinical scale for measuringthe duration of post-traumaticammesia. Med J Aust1986;144:596-72.

47 Kelly R. The post-traumatic syndrome: an iatrogenicdisease. Forensic Sci 1975;6:17-24.

48 Panting A, Merry P. The long-term rehabilitation of severehead injuries with particular reference to the need forsocial and medical report for the patient's family.Rehabilitation 1972;38:33-7.49Rosenbaum M, Lipsitz N, Abraham J, Najenson T. Adescription of an intensive treatment project for therehabilitation of severely brain-injured soldiers. Scand JRehabil 1978;1:-1-6.

50 Minderhoud JM, Boelens MEM, Huizenga J, Saan RJ.Treatment of minor head injuries. Clin Neurol Neurosurg1980;82: 127-40.

716

on April 4, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.6.709 on 1 June 1994. D

ownloaded from