52 Abstracts / Physiotherapy 91 (2005) 49–60

A decision algorithm for detection of fall risk status

S.E. Lamb∗, J.M. Guralnik

The Woman’s Health and Aging Study, Warwick Emergency Care and Rehabilitation,

Warwick Medical School, Coventry CV4 7AL, UK

Introduction Preventing falls is a priority in the NationalService Framework for Older People. Recent guidelineshave suggested that older people should be screened annu-ally for risk of falling, but pointed to a lack of evidence toinform the content and format of screening programmes[1].

Aim Decision modelling to inform the content and formatof screening for risk of falling in community dwelling olderwomen.

Methods A population-based sample representing theone-third most disabled women aged >65 years living inBaltimore (USA) (n= 1002). Baseline examination includeda wide range of known risk factors for falling—fall history,fear of falling, self-reported difficulties balancing, chairrise, 4-m walk speed, balance tests, handgrip strength,cognitive status, vision, incontinence and depression.Tree-based regression was used to model the cut pointsand sequence of tests that best predicted fall status over a1- and 2-year follow-up period.

ResultsIn the first year, 18% of women experienced a newfall, 23% experienced a repeat fall, and 16% recovered from

∗ Corresponding author. Tel.: +44 24 7657 4658.E-mail address:s.lamb@wareick.ac.uk (S.E. Lamb).

ci

. B

, Un

0QB

-sushis

theedut-

being a faller. The optimal screening algorithm was twostage with history of previous falls being most important(cut point = one or more falls), and the time to completethe first metre of a short walking test (<1.5 m/s) adding fur-ther significant improvement to the model. The error rate ofthis model was 34%, with most error being associated withpredicting falls in women who had not fallen previously.Longer-term prediction was tenuous.

ConclusionThe results support an algorithm that is simplebut requires face-to-face contact with the older person, andthat should be, at a minimum, annual. Being able to predictpeople who recover from falls is a novel contribution. Mod-els based on self-reported measures, and change in variablesover time will be presented for comparison.

Reference

[1] Lundebjerg N. Guideline for the prevention of falls in older persons.J Am Geriatr Soc 2001;49:664–72.

d:m”b-rcisesed

doi:10.1016/j.physio.2004.09.007

Systematic review of exer

S.D. Liddle∗, G.D

Health and Rehabilitation Sciences Research Institute

Co. Antrim BT37

Introduction Exercise is popularly used in the rehabilitation of chronic low back pain patients; however, a consendoes not exist on the most effective programme design. Treview aimed to investigate current evidence relating totype and quality of exercise being offered within randomiscontrolled trials (RCTs), and to assess how treatment ocomes are being measured.

se and chronic low back pain

axter, J.H. Gracey

iversity of Ulster at Jordanstown, Shore Road, Newtownabbey,

, Northern Ireland

MethodsA two-fold methodological approach was adoptea methodological assessment identified RCTs of “mediuor “high” methodological quality. Exercise quality was susequently assessed according to the predominant exeused within the trial. Outcome measures were analybased on current recommendations[1].

Abstracts / Physiotherapy 91 (2005) 49–60 53

ResultsFifty-four relevant RCTs were identified, of which51 were scored for methodological quality. Sixteen RCTsinvolving 1730 patients qualified for inclusion in this reviewbased upon their methodological quality, and chronicity ofsymptoms; exercise had a positive effect in all 16 trials.Seventy-five percent of programmes (n= 12) incorporatedstrengthening exercises, of which 85% (n= 10) maintainedtheir positive results at follow-up. Supervision and adequatecompliance were common aspects of trials. Seventy-twopercent of trials supervised maintained their positive resultsat follow-up. A wide variety of outcome measures wereused. Outcome measures did not adequately represent theguidelines for impairment, activity and participation, andimpairment measures were over-represented at the expenseof others[2].

ConclusionsThe review highlighted the diversity of exer-cise programmes offered. The key finding is that exercisehas a positive effect on chronic low back pain patients, andresults are largely maintained at follow-up, despite this di-versity. Strengthening is a common component of exercise

∗ Corresponding author.E-mail address:sd.liddle@ulster.uk (S.D. Liddle)doi:10.1016/j.physio.2004.09.008

nd ndr

dfiaryf He

ofan-ms.ery

ndeat

atecol-ies,ith, rarmpl

ncynal

programmes; however, the role of exercise co-interventionsmust not be overlooked. More high-quality trials are neededto accurately assess the role of supervision and follow-up, together with the use of more appropriate outcomemeasures.

Acknowledgment

The authors acknowledge grant support from the Depart-ment of Employment and Learning (NI).

Reference

[1] Bombardier C. Outcome assessments in the evaluation of treatmentof spinal disorders: summary and general recommendations. Spine2000;25:3100–3.

[2] World Health Organisation. International Classification of Functioningand Disability (ICIDH-2). Geneva: WHO; 2000, Full version.

ared

ut ofviceed inwereostse of, butd bys are

6

A national survey of accident awritten advice fo

R. Nakasha,∗, J. Perryb, E.J. PaaWarwick Emergency Care, Centre for Primb Physiotherapy Subject Division, School o

RelevanceLong-term complications occur in about 30%people who suffer acute neck sprain. Effective early magement is believed to be critical to resolution of symptoThis includes promoting positive messages about recovand avoiding the use of soft collars and rest.

Subjects Lead consultants from all 316 UK accident aemergency departments serving catchment areas of grthan 50 000 people.

MethodsPostal survey. Consultants were asked to indicthe use of a range of treatments (including ice, rest andlars, imaging, referral to psychological or physical therapmedications, method of follow up) and the frequency wwhich these treatments were used (all, some, selectedor no cases). They were also requested to provide a saof the written advice material given to patients.

Analysis Survey data were summarised using frequecounts. Two independent reviewers undertook content a

emergency consultant practice aacute neck sprain

elda, M. Williamsa, S.E. Lamba,Health Care Studies, University of Warwick, UKalth and Social Science, Coventry University, UK

,

er

ee

-

ysis of written materials, and key messages were compagainst those recommended by McClune et al.[1].

ResultsThe response rate was 79% (251 responses o316). The most common intervention was verbal adto exercise, reported by 84% of respondents to be usmost or all cases. Other treatments used frequentlywritten advice and anti-inflammatory medication. Mdepartments stated that they advised against the ucollars. Physiotherapy was used in 60% of departmentsfor selected cases only. Written materials were provide34% of consultants. The results of the content analysishown below:

Positive message about recovery (%)

Promote early return to work/activity (%) 5

Advice for/against collar (%) 49/16

Exercise (%) 79

How to contact a solicitor (%) 63