28Accid Emerg Med 1998;15:298-303

Avoidable deficiencies in the delivery of healthcare to children with meningococcal disease

Simon Nadel, Joseph Britto, Robert Booy, Ian Maconochie, Parviz Habibi, Michael Levin

Department ofPaediatrics, ImperialCollege School ofMedicine at St Mary's,Praed Street, London,W2 INY: PaediatricIntensive Care UnitS NadelJ BrittoI MaconochieP HabibiM Levin

Paediatric InfectiousDisease UnitR Booy

Correspondence to:Dr Simon Nadel, Consultantin Paediatric Intensive Care(e-mail: s.nadel@ic.ac.uk).

Accepted for publication28 May 1998

AbstractObjectives-It is apparent that delays andinadequate or inappropriate managementoccur frequently and may contribute tothe continued high mortality seen inmeningococcal disease. An attempt hasbeen made to define the major sources ofdelay or inappropriate treatment.Methods-A prospective, descriptivestudy of children with meningococcal dis-ease referred to a tertiary centre paediat-ric intensive care and infectious diseaseunit. Definitions of optimal care wereestablished at three stages: parental; gen-eral practitioner (GP)/accident and emer-gency (A&E) department; and hospital.Duration of symptoms and managementwere recorded from direct questioning ofparents and carers, and from hospitalrecords.Results-54 consecutive children withmeningococcal disease were recruited tothe study. Delayed parental recognitionoccurred in 16 children. GPs correctlydiagnosed 19 of 35 children. Delay of2.5-21 hours occurred in those who wereincorrectly diagnosed. Two of 15 childrenwho presented to the A&E departmentwith specific features were incorrectlydiagnosed. Hospital treatment was subop-timal in 71%. Shock was not recognised ortreated in 50%, 20% of children hadunnecessary lumbar punctures. Timefrom illness onset to treatment was longerin fatal disease (median 18.3, range 8-24hours), compared with survivors (median12, range 2-48 hours; p<0.01, Mann-Whitney U test).Conclusion-Suboptimal treatment inmeningococcal disease is due to failure ofparents, GPs, and hospital doctors to rec-ognise specific features of the illness.Improvement in outcome could beachieved by public education and bettertraining of clinicians in recognition, re-suscitation, and stabilisation of seriouslyill children.(7Accid Emerg Med 1998;15:298-303)

Keywords: meningococcal disease; avoidable defi-ciences; health care delivery

Meningococcal disease remains an importantcause of childhood morbidity and mortality inthe UK. Each year there are over 2000reported cases, most of which occur in youngchildren.' Despite the availability of potentantibiotics and improvements in paediatriccritical care medicine, the overall mortality for

patients with meningococcal infection remainsat 10%, increasing to 50% for those whopresent with meningococcal septic shock.2Meningococcal infection is characteristically

fulminant and may lead to death in a matter ofhours. Therefore any delay in initiation ofappropriate treatment is likely to have signifi-cant effects on outcome.There are a number of stages in the interac-

tion between a sick child and the health caresystem, at which management decisions maycritically affect the speed with which medicaltreatment is initiated.

Delays in the administration of appropriatemedical care to seriously ill children may beexperienced:

(1) At home, if parents fail to appreciate thesigns of serious illness in their child, or the sig-nificance of the specific signs ofmeningococcalinfection, such as a petechial rash.

(2) Once a parent has recognised that theirchild requires medical attention, difficultiesmay be experienced in consulting with theirdoctor. A general practitioner (GP) may beunable to appreciate the urgency of thesituation because the parents may be unable toconvey the information adequately over the tele-phone. Once the doctor has seen the child, thesigns of serious illness may not be recognised,due either to their doctor's lack of familiaritywith the condition, or to inexperience of junioraccident and emergency (A&E) doctors. Evenif meningococcal disease is recognised, furtherdelays in the administration of antibiotics or intransporting the child to hospital may occur.

(3) Once the child is admitted to hospital,clinicians may be unfamiliar with the signs ofshock or raised intracranial pressure (ICP) inchildren, and inappropriate management ofthese life threatening complications may occur.Furthermore, delays may be experienced ingaining access to an appropriate paediatricintensive care unit (PICU).As a tertiary referral centre for paediatric

infectious diseases and paediatric intensivecare, with a major research interest in menin-gococcal disease, we have participated in thecare of many children with meningococcalinfection. We have noted that unnecessarydelays in administration of appropriate medicalcare, or inappropriate or inadequate manage-ment, both in the community and in hospital,occur frequently, and may contribute to thecontinued high mortality from meningococcaldisease.

AimIn order to evaluate the frequency of delaysand suboptimal management in children withmeningococcal disease, we have carried out a

298

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from

Health care in children with meningococcal disease

prospective study of the first 54 childrenadmitted to a tertiary centre paediatric unitwith this diagnosis between January 1992 andJanuary 1994.

Patients and methodsMeningococcal disease was diagnosed in chil-dren with a typical clinical presentation, that is,a petechial or purpuric rash and fever. Thediagnosis was confirmed by isolation of Neisse-ria meningitidis from blood, cerebrospinal fluid(CSF), or nasopharyngeal culture; or in culturenegative cases, by detection of Gram negativediplococci in CSF, or detection of meningo-coccal capsular polysaccharide in CSF orblood by latex agglutination.

Patients with typical clinical features ofmeningococcal disease, but where microbio-logical investigations were negative were in-cluded ifno alternative aetiology was identifiedafter full microbiological investigation.

Patients were diagnosed with meningococcalsepticaemia if they had fever and a petechial orpurpuric rash and features of systemic infec-tion (tachycardia, tachypnoea, hypothermia, orhyperthermia). Shock was diagnosed by thepresence of hypotension (blood pressure <3rdcentile for age), or in the absence of hypoten-sion, two or more features of inadequate tissueperfusion, including: base deficit >8 mmol/l;capillary refill time >5 sec; core-peripheraltemperature gradient >3'C; oliguria <0.5ml/kg/hour; hypoxia (oxygen saturation <95%in air), or deteriorating neurological status.3

Patients were diagnosed with meningococcalmeningitis if they had fever and a petechial orpurpuric rash with meningism (stiff neck, pho-tophobia, positive Kernig's sign, depressedconsciousness), without the features of shock.As we do not routinely perform lumbarpuncture in the acute phase of illness in a childwith a petechial or purpuric rash, particularly ifthere is any evidence of circulatory insuffi-ciency, we would presume the presence ofmeningitis in the presence of the above clinicalfindings.

In those children who did undergo lumbarpuncture, meningococcal meningitis was con-firmed by the presence of CSF findingsconsistent with bacterial meningitis and bacte-riological confirmation, or in the absence ofpositive culture or rapid antigen results,presumed meningococcal infection as definedabove.

Patients with clinical or laboratory evidenceof both meningitis and septicaemia wereincluded in the septicaemia group.

Information regarding the child's conditionbefore hospitalisation was obtained by directquestioning of parents and carers, and wherepossible from GP referral letters. Details of thechild's condition and management after admis-sion to hospital was gathered prospectivelyfrom the medical and nursing records of thefirst admitting hospital, from our own recordsfor patients who were direct admissions to ourhospital and for tertiary referrals.

DEFINITIONS OF APPROPRIATE MEDICAL CARE

Recognition and management of children withmeningococcal disease was evaluated at thefollowing three stages of health care delivery.

(1) Recognition by parentsThe following features were defined as criteriathat should have enabled parents to recognisethe need to seek medical advice: fever withpetechial or purpuric rash; or fever with at leasttwo of the following signs or symptoms of seri-ous illness: cyanosis; rigors; convulsions; con-fusion, delirium or altered mental state; severeheadache; severe muscle or joint pain; severeweakness (that is inability to walk or stand);persistent vomiting (three or more occasionswithin 12 hours); and cold extremities.The duration of time from appearance of

these symptoms, to when parents first soughtmedical advice from their GP or A&E depart-ment, was determined by direct questioning ofthe parent or guardian as soon as possible afterhospital admission. Where possible both par-ents were questioned, together with other fam-ily members or friends involved in the child'scare. Accuracy of recollection of the child'ssymptoms was confirmed by other members ofthe family, and information derived from GPletters or A&E records and by examiningmedical and nursing records from both thereferral hospital and the PICU.

(2) Recognition and treatment by GPs andA&EdepartmentsFeatures of illness that should have beenrecognised by GPs and A&E officers as requir-ing immediate treatment and referral to hospi-tal or the paediatric service were the same asthose defined above for parents. Appropriateinitial treatment for meningococcal disease byGPs and A&E officers was defined using theDepartment of Health guidelines.4 The firstphysician to see the patient should administerparenteral penicillin if the child has a petechialor purpuric rash with fever, or clinical featuresof meningitis. The child should then be imme-diately referred to hospital. In the absence of apetechial or purpuric rash, but in the presenceof the features of serious illness as defined forparents, the child should have been immedi-ately referred to hospital.The duration of time between medical

consultation and administration of appropriateinitial treatment was noted.

(3) Recognition and management in hospitalClinical criteria for diagnosis ofmeningococcaldisease by paediatricians or other hospitalbased physicians were fever with petechial orpurpuric rash, or in the absence of rash, clinicalsuspicion of sepsis or meningitis. Additionalfeatures which should have enabled cliniciansto recognise impending or actual shock orraised ICP were: (a) shock: see criteria aboveand (b) raised ICP: Glasgow coma scale <8;cardiovascular instability (hypertension andbradycardia); abnormal respiratory pattern orrespiratory depression; asymmetrical, dilated,or sluggish papillary responses; cranial nervepalsy; abnormal posture; and papilloedema.5

299

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from

Nadel, Britto, Booy, et al

Appropriate initial management of shock inthe admitting hospital was defined as adminis-tration of supplemental oxygen treatment orearly, elective intubation and mechanical venti-lation if indicated (that is by impending oractual respiratory, cardiovascular, orneurological failure); volume replacement withcrystalloid or colloidal solutions to correcthypovolaemia (20 ml/kg or more, given rap-idly); avoidance of lumbar puncture, which isrecognised to be contraindicated in patientswith unstable respiratory or cardiovascularparameters.5

Appropriate initial management of raisedICP included avoidance of lumbar puncture;elective tracheal intubation and hyperventila-tion; fluid restriction; and administration ofdiuretic treatment.6

ResultsFifty four children were admitted to the PICUor paediatric infectious disease unit betweenJanuary 1992 and January 1994. There were31 boys and 23 girls, age range 1 week to 15.7years (median 2.95 years). Fifty one werereferred from other hospitals and three wereadmitted directly.

Forty two children had meningococcalsepticaemia and 12 had meningococcal menin-gitis. Fifteen of those with septicaemia also hadclinical or laboratory evidence of meningitis.

In 37 children, meningococcal disease wasconfirmed microbiologically. Ten of the 17children with negative microbiological resultswere treated with parenteral antibiotics beforemicrobiological investigation. Of those withproved infection, 18 had group B and 19 hadgroup C N meningitidis.Time from illness onset to initiation of

parenteral antibiotic treatment ranged from 2to 96 hours (median 18.7 hours). Childrenwith meningitis had longer duration of symp-toms before presentation (median 24, range12-96 hours) compared with those with septi-caemia (median 14, range 2-48 hours; p <0.05, Mann-Whitney U test).

Thirteen children died (24%), all withmeningococcal septicaemia. The mortality ratefor this group was therefore 31% (13/42). Nochild who had meningococcal meningitis with-out features of septicaemia died.

DELAYED OR INAPPROPRIATE TREATMENTUsing the criteria defined above to indicate thetime when parents should have sought medicalattention, delayed presentation to primarymedical services occurred in 16 children(30%). In six cases, parents hesitated to calltheir GP or take the child to A&E during thenight or at weekends. They preferred to wait,relying on repeated doses of antipyretics ortepid sponging and seek medical advice at amore "socially acceptable" time. Eight sets ofparents did not appreciate the significance of anon-blanching rash, thinking the child haddeveloped "insect bites" or "measles".

In all instances of parental delay, parentswere unaware of the signs of serious illness intheir child, such as rigors, persistent vomiting,or alterations in the level of consciousness. In

four cases, worried parents were inappropri-ately reassured by advice given over thetelephone by GPs or A&E staff without thechild being assessed.

Thirty seven children initially presented totheir GP: 35 had signs of meningococcaldisease; 11 (31%) of these were appropriatelytreated with parenteral penicillin and promptlyreferred to hospital.Delayed recognition or suboptimal treat-

ment of meningococcal disease occurred in 24children (65%): eight with typical clinicalfeatures of meningococcal disease were re-ferred to hospital without receiving parenteralpenicillin from the GP; 16 were not recognisedas having meningococcal disease despite thepresence of fever, a petechial or purpuric rash,and clinical features of serious illness asdefined above. These children were neithertreated with antibiotics nor referred to hospi-tal. Their eventual presentation and initiationof treatment was delayed by between 2 and 21hours (median 8.5 hours). Nine children hadrepeated visits to their GP and seven otherswere subsequently taken to the A&E depart-ment without further attempts to see their GP.Two children who initially consulted their

GP had non-specific illness. When they subse-quently developed the typical features ofmeningococcal disease they were appropriatelymanaged.

Seventeen children presented directly to theA&E department: 15 had typical clinicalfeatures of meningococcal disease on presenta-tion. In two of these an incorrect alternativediagnosis was made (idiopathic thrombocyto-penic purpura in one case and rubella in theother), and therefore antibiotic treatment waswithheld. In one other case, although meningo-coccal disease was correctly diagnosed, initia-tion of antibiotic treatment was delayed forfour hours. These three children had between2.5 to 12 hours delay before administration ofappropriate treatment.Twelve (71 %) children had appropriate

treatment in the A&E department. Twochildren presented to A&E with non-specificfeatures of illness.

All children presenting initially to the A&Edepartment were first assessed by a seniorhouse officer in A&E. The two cases where anincorrrect diagnosis was initially made werereviewed by a paediatric registrar. The diagno-sis was corrected when the child's clinical con-dition deteriorated.

All 54 children were eventually admitted tohospital under the care of paediatricians. Sub-optimal management occurred in 29 children(54%). All of these had meningococcal septi-caemia. Failure to recognise shock in childrenwith meningococcal septicaemia occurred in19 of 38 children, resulting in inadequateresuscitation. Lumbar puncture was per-formed in 10 children with clinical features ofshock, and in one child with features of raisedICP. Of these, nine suffered cardiovascular orneurological deterioration after this procedure,including increase in inotrope requirement,

300

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from

Health care in children with meningococcal disease

c0

00

z6

15

10

5

0

* Survivors (n = 29)v Median (range)

12 (2-48) hours**Deaths(n=13)

Median (range)e 18 (2-24) hours

0-5 6-10 11-15 16-20 21-25 >25Delay (hours)

Figure 1 Timefrom illness onset to initiation of treatmentin children with meningococcal septicaemia.

worsening coma, and one child suffered a car-

diorespiratory arrest and subsequent braindeath.Only 13 of 42 children (29%) with meningo-

coccal septicaemia received optimal treatmentafter hospital admission. There was one deathamong this group of children.Among children with meningococcal septi-

caemia, the duration of time from illness onsetuntil initiation of appropriate treatment was

longer in those who died (median 18.3, range8-24 hours), compared with survivors (median12, range 2-48 hours; p < 0.01, Mann-Whitney U test; fig 1).

DiscussionAny study that is undertaken to defineinadequacies in health care delivery to childrenwith meningococcal disease inevitably risksappearing judgmental or critical of the indi-viduals involved in their care. The purpose ofthis study was not to apportion blame, but toattempt to systematically evaluate the fre-quency of delayed diagnosis or suboptimalmanagement in children with meningococcaldisease. This is a necessary prerequisite beforeconsideration of whether improvements inboth public and medical education are re-

quired to achieve an improvement in outcome.We have shown that there are significant

deficiencies of health care delivery to childrenwith meningococcal disease. These deficien-cies consist of delayed recognition of meningo-coccal disease by parents, GPs and hospitaldoctors, together with suboptimal hospitalmanagement. These factors appear to bepresent at all stages of illness progression.One of the major difficulties encountered

during this study was the lack of reliableindicators of the child's condition beforearrival at hospital. We attempted to beobjective by defining the signs and symptomsthat should have led parents to consult medicalattention, as those commonly accepted as indi-cating serious disease in children.! Although itwas sometimes difficult to obtain an accuratedescription of clinical features from parents,who are often extremely distressed, we foundthat most parents had an acute recollection oftheir actions and the child's condition at eachstage of the illness. There was general agree-

ment between events described by parents, orfriends and relatives who may have accompa-nied them.The reasons for parental delay in seeking

mnedical attention are manifold. In most caseswe did not delve deeply into the reasons whythe doctor was not called or the child was nottaken to the A&E department, as it was verydifficult to suggest, at this most difficult time,that the parents should have sought medicalattention earlier. However, delay in seekingmedical attention due to failure of parents toappreciate the severity of illness occurred in 16children. We were struck by how tolerantparents could be of even severe symptoms intheir child. In addition, if medical advice hadbeen sought, either in person or over thetelephone, and was reassuring that the childwas suffering from a "trivial infection", parentswould hesitate to question the doctor's diagno-sis and possibly delay seeking further advicewhen the child's condition subsequently dete-riorated.One of the major difficulties in the early rec-

ognition of meningococcal disease is that earlysymptoms are often indistinguishable fromthose of other common childhood febrileillnesses such as influenza or viral exanthems.8Any public education campaign that aims toencourage parents to seek medical attentionearly in the course of suspected meningococcalinfection inevitably risks burdening GPs andA&E departments with large numbers of wor-ried parents of children with trivial illnesses.However, from our questioning of parents itseems likely that carefully worded, specificadvice on the recognition of serious illness inchildren, and on the recognition of the rash,may help to reduce the duration of time beforechildren with meningococcal infection receivedmedical attention. Such information is readilyavailable from the meningitis charities, theMeningitis Research Foundation and theNational Meningitis Trust. Simple explana-tions as to the significance of a non-blanchingrash (the "tumbler test") and the need forrepeated parental review and action in thepresence of continuing fever, vomiting, orsevere myalgia, which should prompt furtherconsultation with the GP or A&E department,have been shown to improve the time to receiv-ing medical attention (personal communica-tion, Meningitis Research Foundation). Also,clear recommendations from the GP or A&Eofficer about what symptoms or signs to watchfor, and what to do in the event of any deterio-ration in the child's condition, may preventfurther delays, even if initial medical advice wasreassuring. This would then empower the par-ents to take further action, even if it wascontrary to the initial medical opinion. It maybe reasonable for GP surgeries and A&Edepartments to distribute information sheetsproduced by the meningitis charities to wor-ried parents.

It may be difficult to improve parentalrecognition of signs and symptoms of seriousillness in their children. Aggressive publichealth campaigns, or information to parentssuch as the "baby check" have been shown to

301

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from

Nadel, Britto, Booy, et al

improve self referral rates.7 9 A public educa-tion programme was effective in reducing thetime between illness onset and presentationduring a recent epidemic of meningococcaldisease in Norway, where specific informationwas disseminated about how to recognise therash and symptoms of meningococcal disease.9

Delayed recognition has been shown to beimportant in determining outcome of menin-gococcal disease and was confirmed in ourstudy by the relation between duration ofsymptoms before treatment and mortality.9Although the classical clinical features of

meningitis (fever, vomiting, headache, neckstiffness, and photophobia), are recognised bymost practising physicians, there is less aware-ness that the presenting features of meningo-coccal septicaemia are more commonly non-specific (for example fever, rigors, headache,muscle pains, and confusion). These signs maywrongly be interpreted as an indication ofinfluenza or other viral illness.

Since the original study by Steihm andDamrosch in 1966 that evaluated prognosticfactors in meningococcal disease,'0 numerousstudies have confirmed that those children whoare most likely to die of meningococcalinfection are those without meningitis, butwith features of shock, disseminated intravas-cular coagulation, low white cell count inperipheral blood (<10 x 109/1), low erythrocytesedimentation rate (< 10 mm/hour), low Creactive protein (<50 mg/l), and a rapidly pro-gressive purpuric rash."'..4 In this group of chil-dren, non-specific signs and symptoms willrapidly progress to organ failure and shock,which will be fatal in 20%-50% of cases, with-out features of meningitis ever being present. Itmay only be possible to diagnose these childrenas having meningococcal infection once thepetechial rash becomes evident, or whenfeatures of inadequate end organ perfusionappear. Health professionals are thereforefaced with the difficult task of picking out the"needle in a haystack": the one child withmeningococcal infection from the many withtrivial viral illness. Their task may be madeeasier by ensuring that they undress the childin a good light to try to identify the presence ofa rash; by thorough examination, lookingparticularly for poor peripheral perfusion andtachycardia out of proportion to the degree offever; noting the presence of persistent symp-toms such as severe myalgia, often causingimmobility; and severe abdominal pain. In theabsence of these symptoms, clear instructionsto the parents as stated above, or a plannedreview of the child a few hours later may act toreassure parents and doctors.

In assessing initial recognition of meningo-coccal disease by GPs and casualty officers, weused the same criteria which were defined asthose which should have been recognised byparents as indicating serious illness in theirchild. Delayed recognition of meningococcaldisease by GPs occurred in half of the childrenconsulting their GP, despite the presence of apetechial rash and features of serious illness.The main problems in GP recognition thatemerged from our study were: firstly, GPs may

have failed to undress the child or examine arash despite parental concern that a rash waspresent, and secondly, GPs may not haverecognised features of serious illness, andtherefore falsely reassured worried parents,thus making repeated consultation more diffi-cult.There will inevitably be debate as to the reli-

ability of the signs of serious illness in childrenthat we have used in this study. There may alsobe concerns that the parent's description ofevents, often recorded several hours afterhospital admission, will be distorted. Howevereven if methodological problems have resultedin some distortion of the events, there weremany instances where reassurance was givenby GPs or A&E departments, despite apetechial rash or features of serious illnessbeing present.Although not all children had a positive

microbiological diagnosis of meningococcaldisease, the presence of a petechial rash andfeatures of septicaemia or shock as defined arevery specific for meningococcal disease in theUK. This, together with the absence of anyalternative diagnosis and the knowledge thateven without prior antimicrobial treatmentonly approximately 50% of individuals withmeningococcal infection have positive microbi-ology allowed relative certainty of thediagnosis. 5

Initial hospital management of children withmeningococcal septicaemia was suboptimal in70% of our cases. Half of the children withsepticaemia and shock were not recognised asbeing shocked and therefore appropriate treat-ment was delayed. Even if shock was recog-nised, resuscitation was often inadequate.Lumbar puncture in the presence of cardiovas-cular instability or raised ICP iscontraindicated.5 Knowledge of CSF changesin a child with the clinical features of meningo-coccal disease does not contribute to patientmanagement and may be hazardous. This pro-cedure was performed in 10 shocked childrenand one with raised ICP, nine of whomsuffered significant clinical deterioration.

Thirteen children (24%) died from menin-gococcal disease in this study. This is a highermortality rate overall than that which is usuallyquoted.2 However, children referred to a terti-ary centre are necessarily a more severely illgroup. Forty two of 54 children had a diagno-sis of meningococcal septicaemia. While thisgroup of children was more severely ill, thisdoes not bias the results of our study. If thesechildren had been treated more appropriately,they may not have required referral to thePICU.Our finding that only 29% of children with

meningococcal septicaemia received optimalinitial treatment suggests a need for improvedtraining in the recognition and management ofseriously ill children. Recent initiatives such asthe widespread introduction of advanced lifesupport courses (for example advanced paedi-atric life support and paediatric advanced lifesupport) for junior paediatricians should gosome way to addressing some of the issues ofsuboptimal initial hospital management. In

302

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from

Health care in children with meningococcal disease 303

addition, the use of treatment algorithms andclinical guidelines for hospital treatment mayaid in the earlier recognition and institution ofappropriate treatment for the seriously illchild.6 16

ConclusionThis study has shown that delayed or subopti-mal management occurs commonly in childrenwith meningococcal disease. Failings occur inthe delivery of all aspects of health care, fromparental recognition, GP consultation, andhospital management. It cannot be concludedfrom this study whether outcome would havebeen improved if management had beenoptimal in every case. However, there are fewdiseases that progress as rapidly as meningo-coccal septicaemia, or in which relativelysimple therapeutic interventions to correct thedisordered physiology that is present are ascritical to outcome. The impact on mortality ofmeasures to improve public awareness of thedisease and to improve management in thecommunity and in hospital should be evalu-ated.This work was supported by the Meningitis Research Founda-tion.

1 Jones D. Epidemiology of meningococcal disease in Europeand the USA. In: Cartwright K, ed. Meningococcal disease.Chichester: John Wiley, 1995: 147-57.

2 Havens PL, Garland JS, Brook MM, et al. Trends inmortality in children hospitalised with meningococcalinfections, 1957 to 1987. Pediatr Infect Dis J 1989;8:8-1 1.

3 Bone RC. Let's agree on terminology: definitions of sepsis.Crit Care Med 1991;19:973-6.

4 Department of Health and Social Security. Meningococcalinfection: meningitis and septicaemia. London: HMSO, 1988.(CMO(88)2.)

5 Dean JM, Rogers MC, Traytsman RJ. Pathophysiology andclinical management of the intracranial vault. In: RogersMC, ed. Textbook ofpediatric intensive care. Baltimore: Wil-liams and Wilkins, 1992: 639-66.

6 Nadel S, Levin M, Habibi P. Treatment of meningococcaldisease in childhood. In: Cartwright K, ed. Meningococcaldisease. Chichester: John Wiley, 1995: 207-43.

7 Morley CJ, Thonton AJ, Cole TJ, et al. Interpreting thesymptoms and signs of illness in infants. In: David TJ, ed.Recent advances in paediatrics. 9th Ed. Edinburgh: ChurchillLivingstone, 1991: 137-55.

8 Steven N, Wood M. The clinical spectrum ofmeningococcaldisease. In: Cartwright K, ed. Meningococcal disease.Chichester: John Wiley, 1995: 177-205.

9 Gedde-Dahl TW, Hoiby EA, Brandtzaeg P, et al. Somearguments on early hospital admission and treatment ofsuspected meningococcal disease cases. National Instituteof Public Health Annals 1990; 13:45-60.

10 Steihm RS, Damrosch DS. Factors in the prognosis ofmeningococcal infection. J Pediatr 1966;68:457-67.

11 Lewis LS. Prognostic factors in acute meningococcaemia.Arch Dis Child 1979;54:44-8.

12 Leclerc F, Beuscart R, Guillois B, et al. Prognostic factors ofsevere infectious purpura in children. Intensive Care Med1985;1 1:140-3.

13 Sinclair JF, Skeoch CH, Hallworth D. Prognosis ofmeningococcal septicaemia. Lancet 1987;ii:38.

14 Gedde-Dahl TW, Bjark P, H0iby EA, et al. Severity ofmeningococcal disease: assessment by factors and scoresand implications for patient management. Rev Infect Dis1990;1 2:973-91.

15 Cartwright KV, Jones DM. Investigation of meningococcaldisease. J Clin Pathol 1989;42:634-9.

16 Hodgetts T, Brett A, Castle N. The early management ofmeningococcal disease. J Accid Emerg Med 1998; 15:72-6.

on March 12, 2021 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.15.5.298 on 1 Septem

ber 1998. Dow

nloaded from