Original papers [616689]

Original papers
Patient satisfaction surveys asamarket research
toolforgeneral practices
KEVINKHAYAT
BRIANSALTER
SUMMARY
Background. Recentpolicydevelopments, embracing the
notions ofconsumer choice,qualityofcare,andincreased
general practitioner control overpractice budgets have
resulted inanewcompetitive environment inprimary care.
General practitioners mustnowbemoreawareofhow
theirpatients feelabouttheservices theyreceive, and
patientsatisfaction surveys canbeaneffective toolforgen-
eralpractices.
Aim.Asurveywasundertaken toinvestigate theuseofa
patientsatisfaction surveyandwhether aspects ofpatient
satisfaction variedaccording tosociodemographic charac-
teristics suchasage,sex,socialclass,housing tenureand
lengthoftimeineducation.
Method. Asampleof2173adultslivinginMedway District
HealthAuthority weresurveyed bypostalquestionnaire in
September 1991inordertoelicittheirviewsongeneral
practice services.
Results. Levelsofsatisfaction variedwithage,with
younger peoplebeingconsistently lesssatisfied withgen-
eralpractice services thanolderpeople. Women, thosein
socialclasses 1-3N,homeowners andthosewholeft
schoolaged17yearsorolderweremorecriticalofprimary
careservices thanmen,thoseinsocialclasses3M-5,ten-
antsandthosewholeftschoolbeforetheageof17years.
Conclusion. Surveys andanalyses ofthiskind,ifconducted
forasinglepractice, canformthebasisofamarketing
strategy aimedatoptimizing listsize,listcomposition, and
servicequality.Satisfaction surveys canbereadilyincorp-
oratedintomedical auditandfinancial management.
Keywords: patientsatisfaction; surveydesign; general
practitioner services; marketresearch.
Introduction
CIOVERNMENT policy overthepastdecadehasincreasinglyJemphasized theroleplayedbyconsumers inensuring the
deliveryofhighqualityprimaryhealthcare.In1984theGriffiths
reportsuggested thatmonitoring thehealthserviceatthelocal
levelshouldincorporate theopinions andperceptions of
patients.' Todothis,thereportrecommended thattheNational
HealthServiceadoptthemarketresearchpractices widelyused
intheprivatesectorand'responddirectlytothisinformation, act
onitinformulating policy,andmonitorperformance againstit."'
Threeyearslater,spicedwithreferences topatientchoiceand
the'recognition ofthebenefitsforserviceprovision ofgreater
competition,'2 Promoting betterhealthproposed aseriesof
measures whichitdescribed as'leadingthewaytoafamilydoc-
torservicewhichresponds effectively totheneedsofthecon-
sumer'.2 Thepublic's comments weretobe'soughtonlocalser-
KKhayat,BPh,MSc,research officerandBSalter,BA,MA,DPhil,senior
researchfellow,CentreforHealthServicesStudies,University ofKent,
Canterbury.Submitted: 11February 1993;accepted: 30July1993.
CBritishJournalofGeneralPractice, 1994,44,215-219.vicesandtheirviewsonproposed changes fullytakeninto
account' throughpublicopinionsurveysconducted bythethen
familypractitioner committees.2
Withthe1990NHSandcommunity careact,theintroduction
oftheinternalmarketandtheprioritynowbeinggiventoprim-
arycare,therelationship between theconsumer andthegeneral
practitioner hastakenonanimportance beyondthewell-inten-
tionedrhetoricofearlierpolicystatements. Theprogressive shift
togeneralpractitioner fundholding financed through weighted
capitation givesgeneralpractitioners realincentives toinvest-
igatepatients' views.Aspatientsbecome moreaccustomed to
exercising theirchoiceofgeneralpractitioner, sothepressures
ongeneralpractitioners willgrowtounderstand theirmarketbet-
terandtodiscover thereasonsbehindpatientuseandnon-useof
theirservices. Practiceauditandtotalqualitymanagement may
soonbecomepartoftheeveryday language ofgeneralpractice.3
Patientsatisfaction surveys willbepartofthismarketres-
earch.Awell-established areainprimary careresearch, patient
satisfaction surveyswillproveanimportant meansofexploring
consumer demand. Cartwright's studies4'5 firstestablished the
genrewhichothershavesincerefined,bothintermsofthefocus
ofinquiryandthemeasurement techniques employed.6'7
Itwasdecided toinvestigate howfarthesatisfaction survey
couldbeusedasaninstrument fordefining patientgroupsfor
marketing purposes andtoexamine whether patientsatisfaction
withaspectsofgeneralpractice carevariedaccording tosocio-
demographic characteristics.
Method
Thesurveyquestionnaire wasdesigned andpilotedattheCentre
forHealthServices Studies, University ofKent.Respondents
wereaskedabouttheiruseofgeneralpractice services and
accesstothem,aspectsofgeneralpractitioner care,compliance
withmedication andknowledge.
Fivepointordinalscaleswereusedforanswers tomostques-
tions,from'verysatisfied' to'verydissatisfied' orfrom'agree
strongly' to'disagree strongly'. Severalaskedpatientsaboutsat-
isfaction generally andinthelastconsultation withthedoctor.
Theelectoral registerwastakenasthesampling frameforthe
adultpopulation (aged17yearsandover)ineachofthe53wards
inMedway DistrictHealthAuthority. Systematic random sam-
plingdrewonein80namestocomprise atotalsampleof3146
people.Thiswasslightlyoverthetargetof3000,considered to
besufficient forarangeofmultivariate analyses, theprecise
natureofwhichwasunknown atthetimeofsampling. Eachper-
sonwassentaquestionnaire forself-completion inearly
September 1991.Areminder postcard wassenttonon-respond-
entsafterapproximately threeweeks,followed byaletterand
secondquestionnaire threeweekslater.
Allthequestionnaires wereexamined beforecodingandthe
codingitselfwasrechecked forerrors.Theopen-ended questions
werewrittenintoaword-processing programme andreviewed,
thencodedaswell.AlldatawereinputintotheSPSSstatistical
package. Someoftheinitialfrequencies, suchasthetimeinmin-
utestogettothesurgery, weregrouped. Singleyearsofagewere
alsogrouped. Information onemployment wastranslated into
socialclassusingtheregistrar general's standard occupational
classification. Theinitialfrequencies werefollowed byaseries
ofcrosstabulations whichwereevaluated usingthechisquare
BritishJournalofGeneralPractice, May1994 215

Original papers
test.Onlydatasignificant attheP<0.05levelarereported.
Reduced samplesizeswereobtained wherequestions applied
toonlythoserespondents whogavecertainanswers toprevious
questions orwhowere,forexample, ofaparticular socialclass.
Thesereduced samples arenotedandtheproportion ofnon-
respondents given.Whenquestions haveonlytworesponses,
however, thosewhodidnotanswerareexcluded inordertoclar-
ifyinterpretation.
Results
Representativeness ofthesample
Of3146questionnaires sent,2173wereeventually returned
(69.1%). Ofthe2173whoreturned questionnaires, 20hadno
doctorand94hadnotseenadoctorwithinthepreceding five
years.Thesepeoplewereexcluded frommuchoftheanalysis,
yielding2153or2079respondents formostquestions. Thesam-
plehadslightlyfeweradultsaged17to44yearsthanthepopula-
tionofMedway (52%versus57%)andslightly moreadults
between theagesof45and74years(42%versus36%)(Medway
DistrictHealthAuthority data,1991).Theproportion ofpeople
aged85yearsandovercloselymatched thatinMedway (1.4%
versus1.5%)andtheproportion ofwomeninthesamplewas
slightlyhigherthaninMedway (56%versus51%).
Satisfaction withaccessandgeneralpractitioner care
Reported overallsatisfaction withaccess,withoverallcare
received fromthedoctorandwithout-of-hours consultations was
high,yetthoseagedbetween 17and44yearswereconsistently
lesssatisfiedthanthoseaged45yearsormore(Table1).Inaddi-
tion,7.1%of1030womencompared with4.8%of798men
weredissatisfied withoverallcarereceived fromthedoctor(X22
4.26,1df,P<0.05).
Consulting thedoctor
Overtwothirds(69.0%) of2153respondents reported thatthey
normally sawthedoctortheythought ofastheirownwhile
26.0%saidtheysawanydoctoravailable. Ofthelattergroup,
52.6%couldarrange toseetheirowndoctor'veryeasily'or
Table1.Satisfaction withaccesstodoctor,overallcarereceived
fromdoctorandout-of-hours consultations, byage.
%ofrespondents aged(years)
%ofall
17-44 45+respondents8
AccesstoGP (n=1022) (n=949) (n=2045)
Verysatisfied/satisfied 75.6 87.2 81.1
Neithersatisfied
nordissatisfied 15.5 9.3 12.7
Dissatisfied/
verydissatisfied 8.9 3.5*** 6.2
OverallcarefromGP(n=1028) (n=956) (n=2059)
Verysatisfied/satisfied 78.7 88.0 83.3
Neithersatisfied
nordissatisfied 13.4 8.9 11.2
Dissatisfied/
verydissatisfied 7.9 3.1*** 5.5
Out-of-hours consultation (n=504) (n=382) (n=917)
Verysatisfied/satisfied 72.4 78.5 74.9
Neithersatisfied
nordissatisfied 19.4 13.4 17.0
Dissatisfied/
verydissatisfied 8.1 8.1 8.1
n=number ofrespondents ingroup.aincluding respondents whodid
notgivetheirage.Chisquareforpeopleaged17-44yearsversus45+
years***P<0.001.'fairlyeasily'iftheywanted to.However, 20.5%couldonlydo
so'notveryeasily' or'withsomedifficulty', 15.1%wereuncer-
tainand11.8% gavenoresponse. Ontheirmostrecentvisitto
thesurgery47.9%of547respondents whodidnotnormally see
thedoctortheythoughtofastheirownmanaged toseetheirown
generalpractitioner while52.1% sawsomeone else.Noneofthe
sociodemographic variables revealed significant differences on
thisquestion.
Getting tothesurgery
Fewerthanhalfofthe2079respondents (927,44.6%), attended
thesurgery nearest tothem,36.9%saidtheirs wasnotthenear-
est,15.8%saiditwasaboutthesamedistance asothersurgeries
and2.7%didnotknowordidnotanswer.Thoseinhighersocial
classes weremorelikelytoreportthattherewereothersurgeries
aboutthesamedistance fromhome:18.9%of513respondents in
socialclasses 1and2compared with11.5%of304respondents
insocialclasses4and5described theirsurgeryinthisway(X2=
7.70,1df,P<0.01). Conversely, socialclasswasnotasignifi-
cantvariable amongthosewhoreported thattheirdoctor's
surgery waseitherthenearest ornotthenearest. There were
sharpdivisions according tosocialclassandhousing tenurein
themodeoftransport tothesurgeryandthetimetakentoreach
thesurgery, yetreported easeofreaching thesurgery wasonly
significant forhousing tenure(Table2).
Waitingtimesatthesurgery
Themajority ofrespondents reported normally waiting for
between 10and30minutesfortheirappointment (1382,66.5%).
Thesamplereported shorterwaitingtimesonthelasttriptothe
surgery:65.5%of1951respondents waitedlessthan20minutes,
and26.8% sawthedoctorwithin10minutes. Approximately
equalnumbers waited45minutes ormoreonthelastoccasion
andnormally (6.9%of1951and7.2%of1965,respectively). In
response tothequestion 'Isthewaitingtimetoolong?'39.6%of
1018respondents agedbetween 17and44yearsanswered yes
compared with23.6%of939respondents aged45yearsormore
(x2=57.13,1df,P<0.001).
Contacting thedoctor outofhours
Of959respondents tothequestion, 46.5%reported havingtried
tocontacttheirdoctoroutofhours.Theserespondents weresig-
nificantly morelikelytobemarried(P<0.001), agedbetween 25
and44years(P<0.001), insocialclasses 1and2(P<0.05) and
tobewomen(P<0.01) thanthosenotreporting havingtriedto
contacttheirdoctoroutofhours.
Generally, 39.8%of959respondents reported beingableto
reachtheirowndoctor orapartner outofhours,29.8% werenorm-
allyseenbyanotherdoctor,8.8%received advicebytelephone,
18.1%didnotknow ordidnotanswer,and3.4%indicated that
something otherthantheseoutcomes happened. Onthemost
recentattempt tocontactthegeneralpractitioner outofhours
35.8%of959respondents reported seeingtheirowndoctor ora
partner,37.7% wereseenbysomeone else,17.5%received tele-
phoneadvice,4.2%didnotknow ordidnotanswerand4.8%
reported 'other'.
Onlysocialclasswassignificantly associated withlevelsof
satisfaction without-of-hours consultations: 71.2%ofthe417
respondents insocialclasses1-3Nweresatisfied versus79.3%
ofthe363respondents insocialclasses3M-5(X2=6.82,1df,
P<0.01).
Relationship withthegeneralpractitioner
Inanswer tothequestion 'Doyoulikeyourgeneralpractitioner
asaperson?' 86.0%of2079respondents tickedeitheryesoryes
BritishJournal ofGeneral Practice, May1994KKhayat,BSalter
216

KKhayatandBSalter Original papers
Table2.Detailsofgettingtothesurgery, bysocialclassandhousing tenure.
%ofrespondents
Insocial Insocial Whoare Whoare
classes1-3N classes3M-5 homeowners tenants
Modeoftransport tosurgery (n=835) (n=742) (n=1361) (n=186)
Walking 30.9 40.0 32.4 54.3
Publictransport 2.5 5.5 3.7 17.7
Car 66.6 54.4** 63.9 28.0**
Timetaketoreachsurgery(minutes) (n=971) (n=847) (n=1577) (n=(228)<5 51.6 46.4 49.5 39.0
5-9 29.0 27.7 29.2 18.9
10-14 11.7 13.9 12.9 16.2
15+ 7.6 11.9** 8.4 25.9***
Easeofreaching surgery (n=919) (n=805) (n=1493) (n=213)
Very/fairly easy 92.4 90.9 91.8 84.5
Neither easynordifficult 5.2 6.5 5.8 9.9
Ratherdifficult/very difficult 2.4 2.6 2.5 5.6*.
n=number ofrespondents ingroup.Chisquareforthoseinsocialclasses1-3Nversus3M-5,orhomeownersversustenants**P<0.01, ***P<0.001.
verymuch,10.7%didnotliketheirdoctorand3.3%gaveno
answer.Thenumberofrespondents whoreported likingtheir
doctor'verymuch'increased steadilywithage:19.8%of212
peopleaged17-24years,23.3%of787aged25-44years,30.2%
of573aged45-64years,33.5%of230aged65-74yearsand
42.2%of135aged75+years.Moretenantslikedtheirdoctors
'verymuch'thandidhomeowners(39.3%of211versus26.5%
of1463:x2=14.98,1df,P<0.001).
Ofthe994respondents aged45yearsandover,62.0%felt
theycoulddiscusspersonal problems withtheirdoctorswhile
38.0%felttheycouldnot.Ofthe922respondents agedlessthan
45years49.6%felttheycoulddiscusspersonalproblems with
theirdoctorand50.4%felttheycouldtalkonlyaboutmedical
problems(X2=29.99,1df,P<0.001). Tenants weremorelikely
toreportfeelingabletodiscusspersonal problems withtheir
doctorthanwerehomeowners(62.9%of213versus55.0%of
1444,respectively) whilejustoverathirdoftenants(37.1%of
1444)but45.0%of1444ownersfeltabletodiscussonlymed-
icalproblems(X2=4.73,1df,P<0.05).
Satisfaction withthedoctorintheconsultation
Severalquestions wereaskedaboutspecificaspectsofthecon-
sultation, bothgenerally andonthelastoccasion andtheresults,
showing theoverallsamplefrequencies andthoseofyounger and
olderrespondents, areshowninTable3.Thoughsatisfaction was
high,moreofthoseagedbetween 17and44yearswerelikelyto
belesssatisfiedthantheolderrespondents.
Prescriptions andcompliance withmedication
Mostrespondents feltthattheirdoctorwasreasonable inpre-
scribing (80.1%of2079).Approximately equalnumbers
believed theirdoctors weretooinclined togiveprescriptions
(7.3%),reluctant (4.6%), orwereuncertain (5.6%).Moreof
thoseaged17-44yearsfelttheirdoctorsweretooinclined to
prescribe thanrespondents aged45yearsormore(10.4%of
1011versus4.5%of946,X2=23.85,1df,P<0.001). Younger
peoplewerelesswilling totakemedication andtofinishthe
course,bothgenerally andonthelastoccasion: 63.4%ofthe
1023respondents aged17-44yearsreported thattheygenerally
alwaystookprescribed medication compared with79.7%ofthe
955respondents aged45yearsormore.Atotalof95.1%ofthe
75617-44yearoldsreportedtakingtheirprescribed medication
onthelastoccasion compared with99.0%ofthe79545+year
olds.Of1022respondents aged17-44years56.8%reported thattheygenerally alwayscompleted acourseofmedication com-
paredwith77.5%ofthe944respondents aged45+years;87.9%
of75617-44yearoldscompleted theircourseofmedication last
timecompared with95.6%of79545+yearolds.
Knowledge andabilityofpatients
Askedhowstronglytheyagreedordisagreed withthestatement
'Patients areableandknowledgeable enoughtojudgethetech-
nical/medical skillsoftheirdoctor',38.4%of2079respondents
agreedstrongly oralittle,23.2%disagreed strongly oralittle,
and35.7%neitheragreednordisagreed. Thequestion wasnot
answered by2.7%.Thoseaged45yearsandoverweremore
likelythantheyounger respondents toagreewiththestatement
(42.4%of934versus36.6%of1016;X2=6.81,1df,P<0.01)
andlesslikelytodisagree(19.9%of934versus27.8%of1016;
x2=16.41,1df,P<0.001). Aninteresting patternemergedaccording tohousing, socialclassandeducation variables. There
wasnostatistical difference intheproportions whoagreedwith
thestatement. Oftheremainder, homeowners, thoseinsocial
classes1-3Nandthosewholeftschoolaged17yearsormore
weremorelikelytodisagree withthestatement whiletenants,
thoseinsocialclasses3M-5andthosewholeftschoolaged16
yearsorlessweremoreinclined tochoosetheneutralresponse
(Table4).
Discussion
Fivepointordinalscaleswereusedforrespondents tograde
answersastheseareconsidered tobetterreflectrespondents' true
feelingsthanyes/nooragree/disagree andtherefore increasereli-
ability.8Previous surveyshavetendedtoeitheraskpatientsabout
satisfaction generally orwiththelastconsultation; thisstudy
askedaboutboth.8
Theresultsofthisdistrict-wide surveyareinlinewithmost
localandnational evidence. Highlevelsofoverallsatisfaction
withbothaccesstoandcarereceived fromthedoctorhavebeen
reported byresearchers usingvirtually thesameinstrument in
anotherKentdistrict.9"0 Theroyalcommission ontheNHSalso
confirmed highratesofsatisfaction withaccess.1'
Somelowlevelsofsatisfaction withspecific aspectsofthe
consultation werereported, forexample 24%ofrespondents
reported thatthedoctorgavetheminsufficient information in
theirlastconsultation. Levelsofdissatisfaction withdoctors'
explanations matchedthosefoundbyCartwright4 andCartwright
andAnderson;5 dissatisfaction withtheinformation received
BritishJournalofGeneralPractice, May1994 217

KKhayatandBSalter Original papers
Table3.Satisfaction withthedoctorintheconsultation bothgenerally andonthelastoccasion, byage.
%ofrespondents reporting %ofrespondents reporting
consultations generally lastconsutation
aged(years) aged(years)
%ofall %ofall
17-44 45+ respondentsa 17-44 45+ respondentsa
IsyourGPunderstanding? (n=1002) (n=944) (n=2021) (n=997) (n=925) (n=1991)
Very/fairly understanding 81.6 90.6 86.2 78.7 88.4 83.5
Neitherunderstanding norlacking
inunderstanding 9.9 4.8 7.3 16.0 5.3 8.2
Notvery/totally lackingunderstanding 8.5 4.7*** 6.5 10.2 6.3*** 8.3
IsyourGPgoodatexplaining things?b (n=1018) (n=944) (n=2035)
Verygood/good 65.1 77.8 71.3 – – –
Neithergoodnorbad 27.0 18.3 23.0 – – –
Bad/very bad 7.9 3.9*** 5.7 – –
DoesyourGPgiveyouenough
information? (n=1000) (n=925) (n=2000) (n=1003) (n=922) (n=1998)
Fartoomuch/slightly toomuch 0.6 0.6 0.6 1.4 0.3 0.9
Aboutright 75.4 83.2 79.4 70.8 79.7 75.2
Slightly toolittle/far toolittle 24.0 16.1*** 20.0 27.8 20.0*** 23.9
DoesyourGPspendenough timein
theconsultation? (n=1018) (n=949) (n=2041) (n=1014) (n=946) (n=2033)
Yes 74.0 83.1 78.8 81.9 87.3 84.6
No 26.0 16.9*** 21.2 18.1 12.7*** 15.4
DoesyourGPexamine youthoroughly
whennecessary? (n=1006) (n=943) (n=2023) (n=998) (n=931) (n=1999)
Yes 74.8 79.9 77.3 79.5 88.2 83.7
No 25.2 20.1** 22.7 20.5 11.8*** 16.3
n=number ofrespondents ingroup.alncluding respondents whodidnotgivetheirage.bRespondents notaskedwhether GPwasgoodatexplaining
thingsatthelastconsultation. Chisquareforpeopleaged17-44yearsversus45+years**P<0.01, ***P<0.001.
fromdoctorswasslightlylowerthaninthestudybyWilliams
andCalnan9 butwithintherangefoundbyKinceyandcol-
leagues.'2 Waitingtimeswereslightlylowerthanthosereported
elsewhere 5,12-15butdissatisfaction wasslightlyhigher.5" 2
Cartwright andAnderson reported thatpatients feltmore
knowledgeable aboutprimary carein1977thantheyhad10
yearsearlier,anddoctorsagreed.5Overathirdofrespondents in
thisstudy(38%)feltpatients wereknowledgeable enough to
judgetheirdoctorsandnearlyasmany(36%)wereunsureabout
howknowledgeable patients werewhile23%feltpatients were
notabletojudgetheirdoctors' skills.Analysis bysociodemo-
graphicvariables revealed divisions between thenegative and
neutralresponses andthismaybebecausehomeowners,those
wholeftschoollaterandthoseinsocialclasses1-3Nweremore
certainormoreconfident intheiropinions or,although unprov-
able,becausethesegroupsweremorelikelytodefertotheirdoc-
tors'expertise.Thoseagedbetween 17and44yearswerelesslikelytolike
theirdoctors, werelesssatisfied withthedoctorandweremore
willingtoquestion thedoctor's competence thanolderrespond-
ents,findingssupported byprevioussurveys.4'5'9"6 Ofcourse,the
dataindicatethat,overall,mostpeopleweresatisfied, regardless
ofsociodemographic variables. However, thesevariables are
important indistinguishing betweenrelativelevelsofsatisfaction
andforidentifying dissatisfied patients. Generally, analystshave
notanalysed sociodemographic variables orhavefoundfewdif-
ferences'6 andhavenotconcentrated ondissatisfaction.'7 These
oversights undermine theutilityofresearch onpatientsatisfac-
tionwhichshouldbeundertaken precisely todiscernwhich
patientsaredissatisfied withwhichservices. Anearlystudycon-
sideredthedifference withinapracticebetweenpatientsreferred
tohospitals andpatientsnotreferred. Whilesatisfaction withthe
generalpractitioner's communication rangedbetween 61%and
79%fordifferent sortsofinformation, thepercentage ofpatients
Table4.Disagreement withorneutralresponses tothestatement 'Patients areableandknowledgeable enough tojudgethetechnical/
medical skillsoftheirdoctor',bytenure,socialclassandagewhenleftschool.
%ofrespondents who
Areinsocialclass Leftschoolaged(years)
Arehomeowners Aretenants 1-3N 3M-5 17+ 416
(n=897) (n=126) (n=545) (n=490) (n=337) (n=812)
Neitheragreenordisagree
withstatement 59.1 71.4 49.7 70.2 50.1 64.3
Disagree wtihstatement 40.9 28.6** 50.3 29.8*** 49.9 35.7***
n=number ofrespondents ingroup.Chisquareforhomeowners versustenants, thoseinsocialclasses1-3Nversus3M-5,orthoseleavingschool
aged17+or416years**P<0.01, ***P<0.001.
BritishJournalofGeneralPractice, May1994218

KKhayatandBSalter Original papers
notreferred tohospital whoweresatisfied wasalmosttwicethe
percentage ofpatients whowerereferred whoweresatisfied."2
Thepresentstudyfoundthesameproportion (38%of45+year
olds)asinaprevious study9whofeltthattheycouldnotdiscuss
personal problems withtheirdoctorbutthepresentstudyalso
foundthathalfofthoseaged17-44yearsand45%ofhomeown-
ersfeltthisway.
Inthecontextofmarketing services, itistheserelativediffer-
enceswhichwillmostaffectpractice strategies forimproving
andmaintaining highlevelsofaccess,patientcompliance and
qualityofconsultations. Forinstance, waiting timeshave
received politicalattention andhavebeenshowntoaffectpatient
satisfaction.5'11'12"15 Respondents tothissurveyreported shorter
waitingtimesonthelastconsultation thangenerally, whichmay
indicate atrendtobettertimemanagement withinpractices, or
theeffectsofpatientrecall.The17-44yearoldsweremore
likelytobedissatisfied withwaitingtimesthanthe45+year
olds.Younger peoplereportedbeingmoresatisfied withthedoc-
torinconsultation onthelastoccasion thangenerally butwere
stillsignificantly lesssatisfied thanolderrespondents. Thisevid-
encearguesinfavourofdisaggregated analysis regardless of
whetherthesituation appearstobeimproving overall.
Thestandard satisfaction surveythushasthepotential tohelp
accomplish severalgoalswhichcomplement thepresentchanges
ingeneralpractice. First,itprovides acommunity oratleasta
list-based elementinmedicalaudit.Anobviousexample iswait-
ingtimes,butdataoncompliance andlevelsofsatisfaction for
different aspectsofcareareallimportant. Onestudyfoundthata
composite satisfaction scorewaspositively correlated with
patientcompliance withadvicefromthegeneralpractitioner. 12In
thepresentstudythe17-44yearoldswerebothlesscompliant
withmedication andlesssatisfied withoverallcarethanthe45+
yearolds.Ifsatisfaction influences compliance, andbettercom-
pliance meanshealthier (andlesscostly)patients inthelong
term,thenperhapsthemosteffective waytoimprovecompliance
foryoungerpatientsistoincreasetheirgeneralsatisfaction with
thepractice. Incorporating surveysintoauditmovestowardful-
fillingthesuggestion fromtheGriffiths reportthatsuchinforma-
tionbegathered andincreases thechancesthattheinformation is
usedtoimproveservices.1
Secondly, itishowtheinformation isinterpreted andused
whichunderlies howsurveyscanbemostvaluable toindividual
practices. Surveysidentifypatientgroupswhoaremoreorless
satisfied andthusmoreorlessvulnerable tobeingluredelse-
where.Consider transport anddistance: differences werefound
inmodeoftravelandtimetakentoreachthesurgeryaccording
tosocialclassandhousing tenure.Astudycomparing apractice
inadepressed areawithanaffluentareapracticefoundnodiffer-
encesintheamountoftimetakentogettothesurgeryaccording
toagebutdidfindthatpatientsattending thepractice inthe
depressed areaweremorelikelytowalkwhilethoseattending
theaffluent areapractice werelikelytodrive,withequaluseof
publictransport bybothsetsofpatients.'4 Clearlythereisvari-
ationbetween areasandbetweenpractices. Thefindingthatmore
respondents insocialclasses1or2werelikelytoreportthat
therewereothersurgeries aboutthesamedistance fromhome
thanthoseinlowersocialclassesmaysuggestthatthosein
highersocialclasseshavegreaterchoiceofsurgery.
Identifying whichpatients arelesssatisfiedwithwhichaspects
ofthepracticehelpstargetlistsizeandtherefore generalpracti-
tionerremuneration. According tothe1990contractforgeneral
practitioners, attracting andkeepingpatients aredeemedworthy
offinancialreward.'8 Fromamarketing standpoint, itisimport-
antthatgeneralpractitioners areabletoenhance thosepractice
characteristics mostlikelytoretainexistingpatientswhohave
onlyamarginal commitment ontheonehandandtointerestpotential newpatients ontheother.Questions suchaswhether
out-of-hours consultations shouldbeexpanded, curtailed or
revamped,'9 whether morecounselling shouldbeprovided, and
whether information leafletsareeffectively usedorusedatall
arejustsomeexamples oftheservicewhichmaybeadjusted in
ordertomakethepractice moreattractive.
Finally,itisfurtherpossible thatdatafrompatientsurveysbe
combined withevidence ongroupsatrisk,socialdeprivation,
anddiseasepatternstoaddanotherdimension toepidemiological
evidence aboutpractices andcommunities. Apaperhasexplored
variations inout-of-hours consultations foronegeneralpractice
byelectoral ward-based deprivation scores.20 Inthiscontext, sur-
veyscontribute asmuchtoplanning services asproviding
information onhowtomarketthem:thepotential hereisinneed
ofexploration.
References
1.Department ofHealthandSocialSecurity. NHSmanagement
inquiry.London:HMSO,1984:9.
2.Department ofHealthandSocialSecurity. Promoting betterhealth.
London:HMSO,1987:12,18,61.
3.BakerR.Auditandstandards innewgeneralpractice. BMJ1991;
303:32-34.
4.Cartwright A.Patients andtheirdoctors. London:Routledge and
KeganPaul,1967.
5.Cartwright A,Anderson R.Generalpracticerevisted. London:
Tavistock, 1981.
6.Fitzpatrick R.Satisfaction withhealthcare.In:Fitzpatrick R,Hinton
J,Newman S,etal(eds).Theexperience ofillness.London:
Tavistock, 1984.
7.JonesL,Leneman L,Maclean U,etal.Consumerfeedback forthe
NHS.London: KingEdwards's Hospital FundforLondon, 1987.
8.Fitpatrick R.Surveysofpatientsatisfaction II.Designing a
questionnaire andconducting asurvey.BMJ1991;302:1129-1132.
9.Williams SJ,CalnanM.Keydeterminants ofconsumer satisfaction
withgeneralpractice. FamPract1991;8:237-242.
10.CoyleJ,CalnanC,Williams S.Changing perceptions. Nursing Times
1993;25:44-46.
11.RoyalCommission ontheNational HealthService.Accessto
primary care.Research paper6.London: HMSO,1979.
12.KinceyJ,Bradshaw P,LeyP.Patients' satisfaction andreported
acceptance ofadviceingeneralpractice. JRCollGenPract1975;
25:558-566.
13.RitchieJ,JacobyA,BoneM.Accesstoprimaryhealthcare.
London:HMSO,1981.
14.WykeS,Campbell G,MacIver S.Provision of,andpatient
satisfaction with,primarycareservicesinarelatively affluentarea
andarelatively deprived areaofGlasgow. BrJGenPract1992;42:
271-275.
15.Atkinson C,Gardener M.Opentosuggestions. HealthServJ1991;
101:29.
16.Fitzpatrick R,Hopkins A.Problems intheconceptual framework of
patientsatisfaction research: anempirical exploration. SociolHealth
Illness1983;5:297-311.
17.Carr-Hill R.Themeasurement ofpatientsatisfaction. JPublicHealth
Med1992;14:236-249.18.Department ofHealthandtheWelshOffice.Generalpractice inthe
National HealthService: anewcontract. London:HMSO,1989.
19.KhayatK,WallB.Patientsatisfaction withprimary care.
Canterbury: CentreforHealthServicesStudies,University ofKent,
1992.
20.CarlisleRD,Johnstone SP,PearsonJCG.Relation betweennight
visitratesanddeprivation measures inonegeneralpractice. BMJ
1993;306:1383-1385.
Acknowledgements
Funding fortheprojectwasfromMedway DistrictHealthAuthority and
theKentFamilyHealthServices Authority. Thequestionnaire was
designed andpilotedbyaresearch teamledbyProfessor MichaelCalnan.
Address forcorrespondence
KKhayat,GeorgeAllenWing,University ofKent,Canterbury, KentCT27NF.
BritishJournalofGeneralPractice, May1994 219