*

edaOirtabycpmffwsvLswpaapaoowLpt

RA

The Journal of Emergency Medicine, Vol. 32, No. 1, pp. 15–18, 2007Copyright © 2007 Elsevier Inc.

Printed in the USA. All rights reserved0736-4679/07 $–see front matter

doi:10.1016/j.jemermed.2006.05.032

OriginalContributions

UTILITY OF LUMBAR PUNCTURE IN THE AFEBRILE VS. FEBRILE ELDERLYPATIENT WITH ALTERED MENTAL STATUS: A PILOT STUDY

Kaushal Shah, MD,* Kathleen Richard,† and Jonathan A. Edlow, MD‡

Department of Emergency Medicine, St. Luke’s-Roosevelt Hospital, New York, New York, †Dartmouth Medical School, Hanover, NewHampshire, and ‡Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts

Reprint Address: Kaushal Shah, MD, St. Luke’s-Roosevelt Hospital, 1111 Amsterdam Ave., New York, NY 10025

prm

es

Atmifopfcctllaifswow

Abstract—We conducted a pilot study to compare theiagnostic utility of a lumbar puncture (LP) in febrile vs.febrile elderly patients with altered mental status (AMS).ur null hypothesis was that there is no utility of perform-

ng an LP on the afebrile elderly patient with AMS. Aetrospective study was conducted at an urban, universityertiary care referral center. The study population includedll elderly patients (age 65 years and older) who had cere-rospinal fluid (CSF) samples sent to the laboratory over 1ear. A structured chart review was performed. Exclusionriteria were normal mental status, recent neurosurgicalrocedure or presence of a ventricular shunt, and missingedical records. An LP was considered diagnostically use-

ul if it yielded a diagnosis. There were 185 CSF samplesrom elderly patients recorded over 1 year. Sixty samplesere excluded for the following reasons: normal mental

tatus (36), recent neurosurgical procedure (2), presence ofentricular shunt (11), missing medical record (4), repeatP on same admission (7). Of the 125 patients who met the

tudy criteria, 84 patients were afebrile and 41 patientsere febrile. Of the 84 afebrile patients with AMS, 15atients (18%; 95% confidence interval [CI] 10–26%) hadn abnormal LP. Ten (12%) had some form of meningitisnd five (6%) had unclear diagnoses. Of the 41 febrileatients with AMS, 10 patients (24%; 95% CI 11–38%) hadn abnormal LP. Three (7%) had some form of meningitisr encephalitis. Comparing the elderly patient group with-ut fever with the elderly patient group with fever, thereas no statistical difference in the incidence of abnormalPs or diagnostically useful LPs. Based on the results of thisilot study, we were unable to reject the null hypothesis thathere is no utility of performing LP on afebrile elderly

ECEIVED: 7 July 2004; FINAL SUBMISSION RECEIVED: 18 O

CCEPTED: 30 May 2006

15

atients with altered mental status. We would advocate notelying solely on the presence or absence of fever to deter-ine management in the elderly. © 2007 Elsevier Inc.

Keywords—lumbar puncture; elderly; altered mentaltatus; fever; meningitis

INTRODUCTION

ltered mental status is a common reason for presen-ation of elderly patients to the Emergency Depart-ent (ED) (1); in one study, 5% of Emergency Med-

cal Services (EMS) transports were elderly patientsor evaluation of acute cognitive impairment and 75%f those patients were admitted to the hospital (2). Theotential etiologies were numerous; however, an in-ection is often the source, especially in cases where theause of the altered mental status is not immediatelylear. Frequently, regardless of temperature level, poten-ial sources of infection are sought, e.g., a urinalysis toook for urinary tract infection and a chest X-ray study toook for pneumonia. If an elderly patient has a fever withltered mental status and there is no clear source ofnfection, performing a lumbar puncture (LP) to assessor central nervous system (CNS) infection seems rea-onable and prudent. However, it is not entirely clearhether cerebrospinal fluid (CSF) should be routinelybtained from elderly patients with altered mental statusithout fever.

2005;

ctober

etsCLw

S

Tww

S

TcsylA

nsla

S

Aoshtoat

vsrmmcde

D

F(anp“rlgdt

D

Stp

Tdwt2iasw

cwfws7

mtay

B

OCdw

16 K. Shah et al.

In this pilot study, we sought to test the null hypoth-sis that there is no utility in performing lumbar punc-ures in the afebrile elderly patient with altered mentaltatus. We wanted to compare the incidence of abnormalSF leukocytosis and percentage of diagnostically usefulPs among elderly patients with altered mental statusho did not have fever vs. those who did have a fever.

METHODS

tudy Design

his was a structured retrospective chart review study. Itas approved by the institutional review board with aaiver of informed consent.

tudy Setting and Population

he study was conducted at an urban, university tertiaryare referral center with 50,000 annual ED visits. Thetudy population included all elderly patients (age 65ears and older) who had CSF samples sent to theaboratory from any location in the hospital betweenugust 16, 2000 and August 15, 2001.Exclusion criteria were normal mental status, recent

eurosurgical procedure or presence of a ventricularhunt, and missing medical records. Data from repeatumbar punctures performed in the same admission werelso not included.

tudy Protocol

ll elderly patients who had CSF sent to the laboratoryver the course of the selected year were reviewed. Totudy the group of patients with altered mental status, wead to exclude those with a normal mental status. Pa-ients with a recent neurosurgical procedure or presencef a ventricular shunt were considered to have a separatend significant reason to evaluate the CSF; therefore,hey were also excluded.

One of the investigators performed a structured re-iew of these charts using a data collection form de-igned in advance. Charts and discharge summaries wereeviewed to extract the following information: docu-ented temperature (without regard to method obtained),ental status (altered or normal), CSF analysis, initial

hest X-ray interpretation, urinalysis results, dischargeiagnosis, and recent neurosurgical procedure or pres-

nce of a ventricular shunt. 3

efinitions

ever was defined as temperature � 38°C or 100.4°Fregardless of method obtained). Elderly was defineds � 65 years of age. Mental status was assumed to beormal unless specifically stated to be abnormal in theatient’s record using terminology such as, “altered,”confused,” “somnolent,” “agitated,” “lethargic,” or “un-esponsive.” An abnormal LP was defined as a CSFeukocytosis � 5 WBC/mm3 in CSF tube #4 (abnormallucose and protein levels were not incorporated in theefinition or reviewed). An LP was considered diagnos-ically useful if it resulted in or led to a diagnosis.

ata Analysis

tandard statistical analyses for mean, standard devia-ion, confidence intervals and Fisher’s exact test wereerformed.

RESULTS

he laboratory recorded 185 CSF samples from el-erly patients over the course of 1 year. Sixty patientsere excluded for the following reasons: normal men-

al status (n � 36), recent neurosurgical procedure (n �), presence of ventricular shunt (n � 11), missing med-cal record (n � 4), and repeat CSF sample on the samedmission (n � 7). There were 125 patients who met thetudy criteria. Of these 125 elderly patients, 84 (67%)ere afebrile and 41 (23%) were febrile.Of the 84 afebrile patients, 15 patients (18%; 95%

onfidence interval [CI] 9.7–26.0%) had an abnormal LPith the CSF white blood cell (WBC) count ranging

rom 5 to 120 and a mean of 31.8 � 40.4. The mean ageas similar for the total 84 patients (77.1 � 8.7) and the

ubgroup of 15 patients with the abnormal LPs (77.5 �.6).

Of the 15 patients with abnormal LPs, 2 had bacterialeningitis, 6 had aseptic meningitis, 2 had lymphoma-

ous meningitis and 5 had unclear diagnoses, even aftern inpatient evaluation. In 10 of the 15 cases, the LPielded a diagnosis.

acterial Meningitis

ne patient had group B streptococcus growth from theSF culture, likely from an epidural abscess diagnoseduring inpatient hospitalization; the CSF had 115 WBCsith 80% neutrophils and the patient’s temperature was

6.6°C. The other patient was on chemotherapy for

lCp

A

T(tfTnb

L

Oliphlmt

U

Tndpmd

(W2(an

Cwge1

Td

sittpmbasfm

weeeuhsp

twb(b“tmt59t

iamreae

Tuctlpts

Utility of LP in Afebrile Elderly Patient 17

ymphoma and developed tuberculous meningitis; theSF had 21 WBCs with 72% neutrophils and theatient’s temperature was 37.7°C.

septic Meningitis

here were 6 patients diagnosed with aseptic meningitisa CSF leukocytosis but a negative CSF bacterial cul-ure). All the patients had a CSF lymphocytosis exceptor one patient who had a predominance of neutrophils.hat patient was diagnosed with sepsis/bacteremia butot bacterial meningitis, presumably because the CSFacterial culture was negative.

ymphomatous Meningitis

ne patient had a history of natural killer cell leukemia/ymphoma and was diagnosed with lymphomatous men-ngitis; the CSF had 11 WBCs with 97% blasts and theatient’s temperature was 36.7°C. The other patient alsoad a significant past medical history, specifically B-cellymphoma, and was also diagnosed with lymphomatous

eningitis; the CSF had 19 WBCs and the patient’semperature was 37.9°C.

nclear Diagnoses

he five patients with abnormal LPs but unclear diag-oses had WBC counts in the range of 5 to 8. Theischarge diagnoses were: acute confusional state, uremia/neumonia, right lower extremity weakness, alteredental status of unclear etiology, and cerebellar syn-

rome with possible paraneoplastic syndrome.Of the 41 patients with fever who had LPs, 10 patients

24%; 95% CI 11–38%) had an abnormal LP with a CSFBC count ranging from 6 to 655 and a mean of 92.4 �

00. The mean age was similar for the total 41 patients76.0 � 7.9) and the subgroup of 10 patients with thebnormal LPs (77.5 � 10.7). Of the 10 patients with ab-ormal LPs, 3 had some form of meningitis/encephalitis.

In this study, the incidence of an abnormal LP (i.e.,SF leukocytosis) among afebrile elderly patients (18%)as not significantly different than that of the febrileroup (24%). Similarly, the incidence of meningitis/ncephalitis was also not significantly different, with2% in the afebrile group and 7% in the febrile group.

DISCUSSION/CLINICAL RELEVANCE

he presence of fever is clearly an important sign in the

iagnosis of meningitis; however, like many other pre- s

entations found in the elderly, a noticeable systemicnflammatory response is not always present. Infection inhe elderly is often occult, and it seems that this may behe case here as well. The results of this retrospectiveilot study suggest that elderly patients with alteredental status without fever can develop meningitis. We

elieve the fact that there were 10 cases of meningitismong 84 afebrile elderly patients with altered mentaltatus is clinically significant. The simple absence ofever should not dissuade the physician from consideringeningitis in the differential diagnosis.This does not mean that all afebrile elderly patients

ithout a clear etiology for their AMS require LP; how-ver, this diagnostic procedure should clearly be consid-red. A lack of fever does not rule out meningitis in thelderly. Surprisingly, elderly patients with fever whonderwent lumbar puncture did not have a significantlyigher incidence of meningitis. The absence of statisticalignificance may be due in part to the low number ofatients in both study groups.

In a study of elderly patients with CNS infection overhe course of 15 years, it was found that 86% of patientsere confused or lethargic and 100% of patients withacterial, tuberculous and fungal meningitis had a feverundefined) (3). Another study of elderly patients withacterial meningitis found only 69% of patients hadimpaired consciousness” and 79% of patients had aemperature � 37.5°C (4). Neither study mentioned howany patients were afebrile and had altered mental sta-

us. In two large studies of adult bacterial meningitis,1–66% of patients had altered mental status and 95–7% had a fever (defined as either greater than or equalo 37.7°C or 38°C) (5,6).

In the four instances of bacterial/lymphomatous men-ngitis, the diagnosis based on LP clearly altered man-gement. It may be argued that the diagnosis of asepticeningitis did not alter management because it only

equires supportive care; however, in the case of anlderly patient with altered mental status, making a di-gnosis may help prevent further potentially invasive andxpensive testing.

LIMITATIONS

his was a retrospective chart review study. We werenable to know the specific thought process used by thelinician in determining the need for LP. It is possiblehat some part of the patient’s presentation was particu-arly suggestive of meningitis and prompted a lumbaruncture, despite the absence of fever. It is possible thathese patients had, for example, a significantly elevatederum leukocytosis, meningismus, or headache with nau-

ea and vomiting that was unexplained. It is not known if

thosto

mrtss

Wntkwi

bibwl

1

2

3

4

5

6

18 K. Shah et al.

hese patients had recently taken antipyretics or had aistory of fever at home. We also did not extract timingf the lumbar puncture with relation to arrival or admis-ion time or with relation to antibiotic administration,herefore we are unable to determine if we missed casesf partially treated meningitis.

A prospective study of elderly patients with alteredental status would eliminate the inherent biases in this

etrospective study and help delineate certain risk factorshat may be identified to reduce the number of unneces-ary LPs performed on the elderly with altered mentaltatus.

CONCLUSIONS

e were unable to reject our null hypothesis that there iso utility of performing an LP on afebrile elderly pa-ients. The incidence of an abnormal LP (i.e., CSF leu-ocytosis) among afebrile elderly patients in this studyas 18% compared to 24% in the febrile group. Approx-

mately 12% of afebrile elderly patients and 7% of fe-

rile elderly patients with altered mental status had men-ngitis/encephalitis. Although this pilot study was limitedy its retrospective design and small sample size, weere able to demonstrate the occasional utility of a

umbar puncture in the afebrile elderly patient.

REFERENCES

. Kanich W, Brady WJ, Huff JS, et al. Altered mental status: evalu-ation and etiology in the ED. Am J Emerg Med 2002;20:613–7.

. Wofford JL, Loehr LR, Schwartz E. Acute cognitive impairment inelderly ED patients: etiologies and outcomes. Am J Emerg Med1996;14:649–53.

. Behrman RE, Meyers BR, Mendelson MH, Sacks HS, HirschmanSZ. Central nervous system infections in the elderly. Arch InternMed 1989;149:1596–9.

. Rasmussen HH, Sorensen HT, Moller-Petersen J, Mortensen FV,Nielsen B. Bacterial meningitis in elderly patients: clinical pictureand course. Age Ageing 1992;21:216–20.

. Durand ML, Calderwood SB, Weber DJ, et al. Acute bacterialmeningitis in adults. A review of 493 episodes. N Engl J Med1993;328:21–8.

. Sigurdardottir B, Bjornsson OM, Jonsdottir KE, Erlendsdottir H,

Gudmundsson S. Acute bacterial meningitis in adults. A 20-yearoverview. Arch Intern Med 1997;157:425–30.