Early detection of health problems in potentially frail community-dwelling older people by general practices - project [G]OLD: design of a longitudinal, quasi-experimental study.

0.31 MB·English

by  Stijnen, Mandy MN

Stijnenetal.BMCGeriatrics2013,13:7
http://www.biomedcentral.com/1471-2318/13/7
STUDY PROTOCOL Open Access
Early detection of health problems in potentially
frail community-dwelling older people by general
practices - project [G]OLD: design of a
longitudinal, quasi-experimental study
Mandy MN Stijnen1*, Inge GP Duimel-Peeters1,2, Maria WJ Jansen3 and Hubertus JM Vrijhoef4,5
Abstract
Background: Due to the ageing of the population, thenumberof frail older people who suffer from multiple,
complex health complaints increases and this ultimately threatens theirability to function independently. Many
interventions for frail older people attempt to prevent or delay functional decline, but they show contradicting
results. Recent studies emphasise the importance of embedding these interventionsinto existing primary care
systems and tailoring care to older people’sneeds and wishes. This article presents the design of an evaluation
study, aiming to investigate the effects and feasibility oftheearly detection of health problems among
community-dwelling older people and their subsequent referral to appropriate care and/or well-beingfacilitiesby
general practices.
Methods/Design:A longitudinal, quasi-experimentalstudy is designed comparing 13 intervention practices with
11 control practices. General practices selecteligiblecommunity-dwellingolder people (≥ 75 years). Practicenurses
from interventionpractices (1) visit older people athome for a comprehensive assessment of their health and
well-being; (2)discussresults with theGP; (3) formulate –if required –a care and treatmentplan together with the
patient; (4)refer patient to care and/or well-beingfacilities;and (5)monitor and coordinate care and follow-up.
Control practices provide usual care and match the intervention practices onthe presence of different primary care
professionals within thepractice. Primary outcome measures are health-relatedquality of life and disability.
Additionally, attitude towards ageing, care satisfaction, health care utilisation,nursinghome admission and mortality
are measured. Some outcomes are assessed by means of a postalquestionnaire (at baseline and after 6, 12, and 18
months follow-up), others through continuous registration over the18-month period. A profoundprocess
evaluation will provide insight into barriers and facilitatorsfor implementing the intervention protocol within
general practices from both the patient and caregiver perspective.
Discussion: The proposed approach requires redesigningcare delivery within general practices for accomplishing
appropriate care for older people. A quasi-experimental design is chosen to closely resemble a real-life situation,
which is desirablefor future implementation after this innovationproves to be successful. Resultsof theeffectand
process evaluationwill become available in 2013.
Trial registration: The Netherlands National Trial Register NTR2737
Keywords: Frailty, Older people, Comprehensive geriatric assessment, Home visit, General practice, Quasi-
experimental design
*Correspondence:[email protected]
1DepartmentofGeneralPractice,SchoolforPublicHealthandPrimaryCare
(CAPHRI),FacultyofHealth,MedicineandLifeSciences,MaastrichtUniversity
MedicalCentre,P.O.Box616,6200MDMaastricht,TheNetherlands
Fulllistofauthorinformationisavailableattheendofthearticle
©2013Stijnenetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.
Stijnenetal.BMCGeriatrics2013,13:7 Page2of10
http://www.biomedcentral.com/1471-2318/13/7
Background their geographical proximity to older people, knowledge
Ageing of the population poses challenges to health of the patient’s medical history, relationship of trust be-
care systems as the number of frail older people who tween doctor and patient, and access to a range of multi-
suffer from complex and/or multiple (chronic) health disciplinary health care and well-being facilities in the
complaints increases [1,2]. A failure to detect health person’s neighbourhood. However, general practitioners
complaints among older people in time may cause un- (GPs)oftendonothaveacompleteoverviewofthehealth
necessary neediness and may threaten their ability to status and functioning of older people [19,20]. A Dutch
functionindependently. study among randomly selected older patients revealed
Strategiescomprisingearlyidentificationofolderpeople that34%ofrecordedhealthproblemsduringahome visit
at risk of poor health and early intervention should pre- were unknown to GPs (mostly psychosocial or physical
vent or postpone the onset of functional decline and complaints, such as depression and urine incontinence)
maintainindependent living[3]. Inthe last decades, there [19]. Similarly, Alessi and co-workers [21] reported that
has been an increased focus worldwide on the develop- three-quarter of the visited older people had at least one
ment of preventive home visitation programmes to sup- major health problem identified that was previously un-
port older people to grow old at home and to prevent or known. This suggests that a comprehensive geriatric as-
delayinstitutionalisation. sessmentinthehomesettingyieldsimportantinformation
There is still an ongoing debate whether these pre- about previously undetected health problems and this
ventive home visits should be part of regular care for might be particularly beneficial for the apparently healthy
older people. Numerous systematic reviews have been olderpeople.
published [4-11], attempting to determine the effective- It is equally important that older people themselves
ness of preventive home visits, but the results remain are aware of their own (unmet) health needs, as this
inconclusive.Discrepancy intheresultsiscaused,among appears to be supportive for maintaining independent
others, bydifferences inthe selection of the target popu- living [22]. It seems that older people tend to discard
lation, intensity and duration of the intervention (i.e., certain health problems or complaints as inevitable as-
number of follow-up visits), or domains included in the pects of ageing, such as in the case of urinary incontin-
multidimensional assessment of older people’s health ence [23], they forget about the occurrence of certain
status [12]. Thus, the question remains which compo- events,suchasinreporting fallingincidents[24],orthey
nents of preventive home visits are effective and for may fail to recognise the significance of symptoms or
which population they are beneficial [13]. Most studies complaints [25]. A multidimensional assessment may
to date employ a randomised design for establishing create awareness of these (unmet) needs or problems.
the success of preventive home visits, thereby hindering After health problems and complaints are identified,
close resemblance to a real-life situation and restricting care facilities should be tailored to older patient’s needs
theexternalvalidity offindings. and preferences [26-28] and active involvement of older
Recent publicationsstressthe importanceofintegrating people in decision-making concerning their need for
preventiveinterventionsforolderpeopleintoexistingcare careservicesisencouraged [29].
systems[10,14,15].Forexample,VanHoutandcolleagues In conclusion, we hypothesise that a multidimensional
[14] attribute the absence of a preventive effect of home comprehensive geriatric assessment of the health and
visits to the fact that they were not integrated within pri- well-being of community-dwelling older people by gen-
mary care practices. In our current approach, instead of eral practices and subsequent individualised care and
solely integrating, we aim to redesign care delivery within follow-up (if required) will lead to improved health-
primary care practices by applying components of the related quality of life and reduced disability compared to
Chronic Care Model (CCM). This comprehensive frame- usual care (i.e., reactive careto olderpeople who present
work has proven to lead to improved patient care and themselves with health problems or complaints). Fur-
betterhealthoutcomeswhenchangingroutinedeliveryof thermore,wehypothesisethatthisapproachwillbefeas-
care through improvements in six interrelated compo- ible from both the patient and caregiver perspective. In
nents(furtherdetailsareprovidedintheMethodssection) the current paper, the design of an evaluation study is
[16,17]. In addition, elements of the Guided Care model presented aiming to investigate the effects of identifying
are incorporated in our approach [18]. Guided Care used health problems and complaints among potentially frail
theCCMtoidentifysuccessfulinnovationsinchroniccare community-dwelling older people at an early stage and,
that can be applied in primary care to achieve optimal if necessary, their follow-up within care and/or well-
outcomes in people with chronic diseases and complex being facilities. Parallel to the effect study, a process
careneeds. evaluation will provide insight into the barriers and faci-
Generalpracticesseemtobetheidealsetting forrealis- litators for implementing the proposed approach within
ing preventive care facilities for older people, because of generalpractices.
Stijnenetal.BMCGeriatrics2013,13:7 Page3of10
http://www.biomedcentral.com/1471-2318/13/7
Methods/Design proactive care). Effects on outcome measures are assessed
Studydesignandsetting at baseline (T ) and after 6-months (T ), 12-months (T ),
0 1 2
Thelongitudinal,quasi-experimentalstudyisperformedin and 18-months (T ) follow-up. Parallel to the effect study,
3
three regions in the south of the Netherlands: Maastricht- aprocessevaluationisperformed.Figure1presentsaflow
Heuvelland (8.5%≥75 years), Parkstad (8.7%≥75 years), chart of the study design and measurements. A more
and Midden-Limburg (7.5%≥75 years). They are particu- complete overview of the study protocol, including a time
larly interesting because the ageing of the population is schedule,isprovidedinFigure2.
more pronounced here (nationwide 7.0%≥75 years). Gen- TheMedicalEthicalCommittee(MEC)oftheMaastricht
eralpracticesintheseregionswereinvitedtoparticipatein University Medical Centre (MUMC+) judged this evalu-
the evaluation study. Participating general practices ran- ation study as not needing formal ethical approval. Ne-
domly selected community-dwelling people aged 75 years vertheless, the MEC granted their approval for our study
and older from the GP Information System. Older people protocolandinformedconsentdocuments.
within intervention practices are visited at home by the
practice nurse for a multidimensional assessment followed Selectionofgeneralpractices
by individualised care, the so-called [G]OLD-protocol: General practices (n=21) who visited older people at
‘Getting OLD the healthy way’. Older people from control home as part of our pilot study [30] were excluded from
practices receive usual care (i.e., reactive care instead of participationtopreventcontaminationofpriorexperience.
General practices
approached for participation, n= 188
General practices willing to participate, n= 24
Included in Included in
intervention group, n= 13 control group, n= 11
Selection older people (age 75 years)
by general practices
h
Applicationexclusioncriteria
Older people eligible for participation
Exclusion of older people due to lack of time of
practice nurses from intervention group to visit all
eligible older people for [G]OLD-consultation within
one-year time
Older people approached for participation
h
Expectedresponserate=50%
Baseline measurement (T)
0
Target n per study group = 858
Intervention: [G]OLD-protocol Usual care Expected
drop-out
rate = 30%
Follow-up: 6 months (T), 12 months (T), 18 months (T)
1 2 3
Analysis
Target n per study group = 600
Figure1Flowchartstudydesignandmeasurements.
Stijnenetal.BMCGeriatrics2013,13:7 Page4of10
http://www.biomedcentral.com/1471-2318/13/7
Selection GP practices for control group
Selection older people (75+)
Information letter and IC to eligible people
Selection
olde(r7 5p+e)ople Control group (C)- T0
[G]OLD care booklet
Information C- T1
Selection
pforraG cintPitceers- leeltitgeirb alen dp eIoCp lteo Intervention group (I) - T0 C- T2 C- T3
[G]OLD care booklet
vention Trai- PNs Trai-
group ning practice ning I- T1
1 [G] 2
PN - OLD- PN - I- T2
Feb. ctoantisounl- May I- T3
[G]OLD-
Eva- Eva- consultation: Formulate
lua- lua- home visit by PN care and
tfioornm tfioornm difmore mnsuioltni-al Post- pltarnea; tdmisecnutss Theoretical framework[G]OLD
assessment discussion with patient
with GP Effect measurement
Components Chronic Care Model (CCM): Additional Referral
-dpreoliavcetriyv es yinstsetema dd eosfi grena: cdteivliev ecrayre o; fmain yes ec(exo.naas.m u[Gltian]tOaioLtniDo 2–n) + foulplow- PMroaicne sesl eemvaelnutast [ioGn]OLD-protocol
task for PN, supported by GP Problems
-decision support: referral based on the that require Activities research team
rgeusiudletlsi noef se,v aindde npcaet-ibenast’esd n teeesdtss /awnidshes attention? Discuss with Follow- Activities GP / PN
-clinical information systems: system for no patient up
registration of assessment results,
development of plan for care/treatment,
monitoring and follow-up
-community resources: establish linkages Interviews PNs (3 per PN) and 2 feedback sessions Interviews
with care disciplines in the neighbour- GPs
hood; collaborate in organising care
Interviews older people
Timeline
Nov. 2009- Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Jan. - Jun. - Dec.
Jan. 2010 2010 2010 2011 2012
Figure2Schematicoverviewstudyprotocol.Note:PN=practicenurse;GP=generalpractitioner;IC=informedconsent.
We approached 188 general practices in the three regions presence of primary care professionals within general
for participation in this study. Practices in Midden- practicestoensure comparabilityatbaseline.Weassume
Limburg were only invited to participate in the control that close proximity of various primary care disciplines
group, since insufficient general practices from the other facilitates collaboration in organising and/or delivering
two regions were willing to participate in the control appropriate care to older people [31]. After the recruit-
group. GPs who indicated to be actively engaged in or are ment phase, one intervention practice and two control
planning to start with the identification and follow-up of practices dropped out due to a lack of time to select
frail older people in a systematic way were ineligible to older people eligible for participation. As a result, 24
participate (n=12 practices). The availability of a practice generalpracticeswere included inthisstudy.
nurse who has time for care for older people is a prere-
quisite for intervention practices. Practice nurses work in Targetpopulation
general practices, and provide screening, treatment, care The target population are the apparently healthy,
andeducationmainlytopatientswithchronicdiseasesand community-dwelling older people aged 75 years and
olderpeople. older. Although the age criterion causes much contro-
Reasons of general practices for non-participation versy, especially from the age of 75 years on the preva-
were: no time (e.g., due to other priorities, staff changes lence of frailty increases markedly [32]. This enables us
or participation in other research projects) (35.0%), no to find sufficient eligible older people for participation.
interest to participate in the present study (31.7%), inter- Furthermore, some authors suggest that preventive home
ested in [G]OLD-intervention but not in research visits are most beneficial for people aged 75 years and
(19.5%),ornoreasonwasmentioned(13.8%). older [33,34]. We excluded people who are not living
Fourteen general practices agreed to participate as independently, those on a waiting list for admission to a
intervention practice and 13 general practices consented nursinghomeorhomeforolderpeople,thoseunderclose
to participate in the control group. Control practices medical supervision (chemotherapy, chronic haemodialy-
were matched to intervention practices based on the sisorothertherapiesposingahighburdenontheperson),
Stijnenetal.BMCGeriatrics2013,13:7 Page5of10
http://www.biomedcentral.com/1471-2318/13/7
and the terminally ill. Practice nurses’ available working visit, the practicenurse makes aprint out ofthe person’s
hours for care for older people determined the maximum medication list and medical history for relevant details
numberofolderpeoplethateachofthemwasabletovisit ormajoreventstobeawareof.
within one year's time. This, together with the size of the During the visit, the practice nurse uses the [G]
patientpopulationaged75yearsandolder,determinedthe OLD-instrument: a structured, comprehensive geriatric
number of older people per intervention practice assessment to assess the person’s physical, psycho-
approached for participation. In control practices, all eli- logical, mental and social functioning, as well as
gibleolderpeopleaged75yearsandoverwereapproached. lifestyle and medication use (see Table 1). This instru-
We invited older people for participation by means of ment is specifically developed for and tested among
an information letter and consent form. We performed the apparently healthy community-dwelling older
telephonic reminders in the intervention group and pos- people aged 75 years and older in a pilot study [30].
tal reminders after four weeks in the control group to Suggestions made during the pilot phase helped to im-
those who didnotrespondtothefirstmailing. prove the [G]OLD-instrument for application in the
current longitudinal, quasi-experimental study. In gen-
Procedure eral, the instrument assists the practice nurse in unco-
Although the independent effects of components of pre- vering (early signs of) potential health problems or
ventive home visitation programmes are difficult to dis- needs that may prevent older people, now or in the
entangle, previous research has suggested elements that near future, from maintaining independent living. Al-
at least should be included, such as a comprehensive though the instrument follows a structured format, it
geriatric assessment, a concrete care plan and multiple can be applied in a flexible way. For each test
follow-up contacts [7,8,29]. We redesigned care delivery included, evidence-based cut-off points and guidelines
for older people by general practices by focussing on are presented to assist in deciding about the presence
several evidence-based elements of the Chronic Care or absence of health problems or needs.
Model (CCM) and the Guided Care model (for details, Crucial during the visit is establishing a relationship of
seeFigure2).Applyingbothmodelshasledtothedevel- trust, listening to the needs and wishes of the older per-
opment of the [G]OLD-protocol, which is explained in son, and allowing the person time to talk [29]. If neces-
more detail below. Our pilot study provided preliminary sary, the practice nurse may also provide information or
evidence of the feasibility of the [G]OLD-protocol for advice. Sometimes it is necessary to perform an add-
generalpractices[30]. itional examination to obtain amore accurateestimation
of the presence of problems. Therefore, more elaborate
Training tests on the themes cognition, depression and personal-
Practice nurses from intervention practices received two ity disorders are incorporated inthe [G]OLD-instrument
days of training before the start of the study to provide part 2 which can be conducted during the first visit or
themwiththenecessaryknowledgeandskillsforexecuting during a second visit, depending on the older person’s
all elements of the [G]OLD-protocol. In this way, we also preference.
attemptedtoequalisethelevelofknowledgeandskillsbe- After the home visit, the practice nurse registers the
tweenpracticenursesregardingcareforolderpeople.Cen- results of the [G]OLD-instrument in the electronic pa-
tral elements of the training included acquiring tient file.
communication skills, gaining knowledge about frequently
occurring health problems among older people, gaining Post-discussionGPandformulatingcareandtreatment
knowledge about health services for older people, and plan
learning how to assess older people’s physical, psycho- Thepracticenursediscussestheresultsofthehome visit
logical,mentalandsocialfunctioningbymeansofamulti- with the GP. The results of the [G]OLD-instrument,
dimensional instrument. In between the two training as well as the patient’s needs and wishes, determine
sessions, each practice nurse performed exactly five home whether follow-up actions regarding certain problems
visits among randomly chosen older people (≥ 75 years) areneeded.Theseactionsmayconsistofadditionaldiag-
duringatry-outphase.Duringtheinterventionperiod,ses- nosis, preventive care or advise, treatment in primary
sions were organised for asking questions and exchanging health care or referral to other care and/or well-being
experiences, and practice nurses received additional sup- facilities as much as possible in the older person’s neigh-
portbyacoachspecialisedingeriatriccare. bourhood. The practice nurse formulates a provisional
individualised care and treatment plan. This plan is
Homevisit-comprehensivegeriatricassessment discussed with the patient, whose input and wishes lead
The practice nurse invites older people for a home visit to a final care and treatment plan, which is registered in
successively within a one-year time period. Before the theelectronic patientfile.
Stijnenetal.BMCGeriatrics2013,13:7 Page6of10
http://www.biomedcentral.com/1471-2318/13/7
Table1Topicsincludedinthe[G]OLDcomprehensivegeriatricassessmentinstrument
Basicassessment–partone Additionalassessment–parttwo
Physicalfunctioningandlifestyle DisabilityinADLandIADL;needforassistanceinADLand/orIADL; N/A
incontinence;mobility;falls;visionandhearingproblems;BMI
(heightandweight);malnutrition;bloodpressure;physicalactivity;
smoking;alcoholuse
Psychologicalfunctioning Cognition;anxiety;depression;personalitydisorders Cognition;depression;personality
disorders
Socialfunctioning Receivingandprovidinginformalcare;loneliness;socialparticipation N/A
Additional Generalperceptionofhealthandqualityoflife;medicationuse;financial N/A
situation;healthcareutilisation;observationoflivingenvironment;
physical,psychologicalandbehaviouralsignals
Note:N/Ameansnotapplicable.
Referralandfollow-up These outcomes, together with the secondary outcome
The practice nurse arranges and coordinates care for the attitude towards ageing (subscale attitude toward ageing
older person as formulated in the final care and treat- from the PGC Morale Scale) [40] are included in a ques-
mentplanandmonitors thefollow-up.Theneed for and tionnaire send to older people by postal mail at baseline,
frequency of follow-up contacts strongly depends on the 6-months, 12-months and 18-months follow-up. The
type of problems or complaints that deserve attention baseline questionnaire also gathers data about socio-
according to the care and treatment plan. Hence, no demographic variables (i.e., age, gender, ethnicity, edu-
fixed number of contacts per older person is determined cational level, marital status, household composition) to
on forehand. The practice nurse indicates in the care provide insight into characteristics of the target popula-
and treatment plan at what date a specific problem or tion. Assistance is provided to older people who are
complaint will be re-evaluated. Then, at each follow-up unable to self-complete the questionnaires or those with
contact, the need for additional follow-up contacts is many missing items (mostly people with poor physical or
determined and, if necessary, the care and treatment mentalhealth).
plan is adjusted. Notably, these follow-up contacts may Additional secondary outcomes are admission to a
also take place with other care providers to whom older nursing home or home for older people, health care
people arebeingreferred. utilisation,andmortality.Generalpracticesregisterthese
If follow-up actions are not required or they are not outcomes continuously during the study period in the
desirable from the patient’s point of view, the practice GP Information System and data are extracted for each
nurse discusses with both the GP and the older person patient after 18-months follow-up. Furthermore, health
how they will proceed from that moment on. It is im- care utilisation is also recorded in the [G]OLD care
portant that general practices should prevent to lose booklet. Older people receive this booklet at baseline
sight of their older patients after this initial assessment. and are requested to take it with them to each contact
The home visit is not meaningless when no specific pro- with professionalhealth careprovidersfor 18months.In
blems are identified, as it helps to gain knowledge about this booklet, patients and/or care providers indicate the
the patient in case future health problems occur (e.g., reason for the contact, type of health problems or com-
falling incidents). Furthermore, the bond created with plaints, and follow-up activities. Table 2 presents all
the practice nurse increases the likelihood that older outcome measures, their operationalisation and timing
people willapproach theirgeneral practiceincaseofany ofdatacollection.
future problems or complaints. Older people who do
not receive follow-up contacts will remain part of the Processevaluation
study population to ensure comparability with the con- A thorough process evaluation is conducted aiming to
trolgroup andtheywillbeanalysedasa sub-group. investigate to what extent the different components of
the [G]OLD-protocol are implemented within general
Measuresanddatacollection practices as planned (e.g., barriers and facilitators for
The primary outcome measures in this study are health- implementation) and the feasibility of the protocol for
related quality of life measured by the RAND-36 [35,36] both patients and caregivers. Ultimately, the results may
and disability in activities of daily living (ADL, including provide information for further implementation of
mobility)andinstrumentalactivitiesofdailyliving(IADL), [G]OLD within general practices. Qualitative and quan-
assessed using the Groningen Activity Restriction Scale titative process data are collected with either formative
(GARS)[37].Bothinstrumentsappeartobevalid,reliable or summative purposes among GPs, practice nurses and
and suitable for self-completion in older people [38,39]. older people according to the comprehensive and
Stijnenetal.BMCGeriatrics2013,13:7 Page7of10
http://www.biomedcentral.com/1471-2318/13/7
Table2Measures,operationalisationandtimingofdatacollection
Measures Operationalisation No.ofitems Rangescore* Timingdatacollection†
Primaryoutcomes
Health-relatedqualityoflife RAND-36[35,36] 36 N/A T,T,T,T
0 1 2 3
Disability GARS[37] 18 18–72 T,T,T,T
0 1 2 3
IADL 11 11-44
ADL 7 7–28
Secondaryoutcomes
Attitudetowardsageing Subscaleattitudetowardownageing-PGCMorale 5 0–5 T,T,T,T
0 1 2 3
Scale[40]
Healthcareutilisation Numberofcontactswithdifferenthealthcare 3 N/A T
0
providers(i.e.,GPconsultations,hospitaladmission)
N/A N/A CR_GPandCR_E
Admissiontonursinghomeor Numberofadmissionsandtimetoadmissionfrom N/A N/A CR_GP
homeforolderpeople T toT
0 3
Mortality NumberofdeathsfromT toT N/A N/A CR_GP
0 3
*Underlinedscoresindicatethemostfavourablescores.N/Ameansnotapplicable.
†T0=postalquestionnaireatthestartofthestudy;T1=postalquestionnaireat6-monthsfollow-up;T2=postalquestionnaireat12-monthsfollow-up;T3=postal
questionnaireat18-monthsfollow-up;CR_GP=continuousregistrationduringstudyinGP’sInformationSystem;CR_E=continuousregistrationduringstudyby
olderpeoplein[G]OLDcarebooklet.
systematic approach proposed by Saunders and colleagues ‘general health perception’) as measured by the RAND-36
[41](fordetails,seeTable3). [35,36].Weaimto demonstrateaclinicallyrelevantdiffer-
The experience of practice nurses with the [G]OLD- ence between the mean change score of the intervention
protocol was assessed three times during one year (inter- andcontrolgroupof5.0onthetransformedsubscale.This
vention period)bymeansofindividualinterviews.Results implies a standardised effect size of 0.24 (givenSD=21.2).
of these interviews were fed back to all practice nurses Based on this and applying a significance level (α) of 0.05
together during feedback sessions after six months and and a power of 0.90, the minimally required number of
after one year (end of the intervention period). Addition- participantsis564(n=282perstudygroup)usinganinde-
ally, GPs were individually interviewed at the end of the pendent samples t-test (two-sided). However, calculations
intervention period to assess their experiences with the thattakeintoaccounttheinterdependencyofthemeasure-
implementation of [G]OLD within their general practice. ments within a cluster (i.e., general practice) and correct
One older person per general practice was selected for for the cluster effect result in a required sample size of
in-depth interviews about their experiences and satisfac- 1,200olderpeople(n=600perstudygroup).
tion with all aspects of the [G]OLD-protocol, approxi- We expected a response rate of 50% on the informa-
mately one month after the [G]OLD-consultation took tion letter and consent form sent to eligible older people
place. Furthermore, older people can register their satis- for participation and a drop-out rate of 30%. Accounting
faction with contacts with professional care givers in the for drop-out, we planned to enrol 858 older people per
[G]OLD-care booklet. Finally, timerequired for the home study group to have a sufficient number of participants
visit,resultsofthetestsperformedandpreliminaryadvise per group (600 older people) at the end of 18 months
giventopeopleduringthehomevisitareregisteredbythe follow-up (see also Figure 1). Because of the expected
practice nurse on the [G]OLD-instrument. Details about response rate of 50%, we planned to approach at least
referral to care and/or well-being services are written 1,716 community-dwelling older people per study group
down in the care and treatment plan. Members of the forparticipation.
research team checked monthly during the intervention Since the amount of home visits that is performed
period to what extent the [G]OLD-instrument and the depends on the PNs available time, we expected a vari-
care and treatment plan were completely filled out. Prac- ationinclustersizes.Thisiscompensatedforbysampling
tice nurses registered the patient’s follow-up within the 25%moreclusters(i.e.,generalpractices)[42].
chainofcareintheelectronicpatientfilefromwhichrele-
vantdatacanbeextractedafter18-monthsfollow-up. Statisticalanalyses
We compute descriptive statistics to describe the charac-
Samplesizeconsiderations teristicsofthetargetpopulationandgeneralpracticesand
The sample size calculation is based on the primary out- to investigate comparability of study groups at baseline.
come measure health-related quality of life (subscale Relevant statistical tests(e.g.,t-test, chi-square,analysis of
Stijnenetal.BMCGeriatrics2013,13:7 Page8of10
http://www.biomedcentral.com/1471-2318/13/7
Table3Datacollectionaspartoftheprocessevaluation Data gathered as part of the process evaluation will be
Components Operationalisation Datacollection analysed using descriptive techniques, such as calculating
Tools/Procedures scores(e.g.,numberofdrop-outs),narrativedescriptionof
Fidelity Extenttowhichthe Evaluationformtraining procedures,andidentifyingthemesintheinterviews.
(quality) [G]OLD-protocolwas PNs
implementedasplanned
Individualinterviews Discussion
PNs,GPsandolder
In the present paper, the design of a longitudinal, quasi-
people
experimental study is presented to investigate the effects
FeedbacksessionsPNs
of the early detection of health problems among com-
[G]OLD-instrumentand munity-dwelling older people and their subsequent fol-
careandtreatmentplan
low-up within the chain of care by general practices. In
Dosedelivered Extenttowhichallaspectsof Evaluationformtraining
contrast to existing studies, we purposefully chose for a
(completeness) [G]OLD-protocolaredelivered PNs
togeneralpracticesand quasi-experimental design. Although randomised con-
olderpeople IndividualinterviewsPNs, trolled trials are widely accepted as the “gold standard”
GPsandolderpeople
for evaluating the effectiveness of interventions, they
FeedbacksessionsPNs
createartificial situations thatmay hinderthe translation
[G]OLD-instrumentand
of research findings into practice [43-45]. Moreover, the
careandtreatmentplan
study may suffer from the uncertain commitment of the
Dosereceived ExtenttowhichPNs,GPsand Individualinterviews
people delivering the intervention (in this case the gen-
(exposure) olderpeopleactivelyengage PNs,GPsandolder
in,interactwithandare people eral practice’s staff) to the changes to be made. Routin-
receptivetoaspectsof isation of working methods in daily practice must take
FeedbacksessionsPNs
[G]OLD-protocol
place to ensure sustainability of the [G]OLD-protocol
[G]OLDcarebookletfor
olderpeople [46], which is more difficult to realise within a rando-
mised design. Our combination of an effect study and a
Dosereceived OverallopinionofPNs,GPs, Evaluationformtraining
(satisfaction) andolderpeopleabout PNs thorough process evaluation should provide sufficient
[G]OLD information with respect to the feasibility and external
Individualinterviews
validity ofthe[G]OLD-protocol within generalpractices.
FeedbacksessionsPNs
Furthermore, we predominantly used the Chronic
[G]OLDcarebookletfor
Care Model for redesigning primary care practice as
olderpeople
applying elements of this framework appears to lead
Reach Proportionofintendedtarget Continuousregistration
to improved patient care and better health outcomes
(participation populationthatparticipates bygeneralpracticesand
rate) inandcompletesthe researchers among patients [17]. We additionally expect that the
intervention: multidimensional [G]OLD-instrument will be of added
(1)registrationnumberand Individualinterviewswith value in providing a comprehensive overview of the
reasonsfornon-responseand PNsandGPs older person’s health status, compared to intervention
drop-out;(2)opinionPNsand
GPsaboutreach programs that only focus on a limited number of tests
orquestionsinonlyone ortwo domains.
Context Environmentalbarriersand Individualinterviewswith
facilitatorsthatinfluence PNsandGPs Challenges faced during the intervention period are
implementation[G]OLD, managing internal and external factors (e.g., changes
Notesresearchersand
continuedinvolvementin[G] projectleaders in the general practice’s policy or reimbursement of me-
OLD,and/orstudyoutcomes
dical expenses by insurance companies) to allow for
Note:PN=practicenurse;GP=generalpractitioner.
continued and adequate implementation. Furthermore,
considering the current interest of general practices in
variance, regression analysis) will be applied to analyse care for older people, general practices who participate
effectsonprimaryandsecondaryoutcomemeasures(level in the control group are closely monitored until the end
of significance is 0.05; two-tailed). Data will be analy- of the follow-up period to find out if they implement
sed according to the intention-to-treat principle. In all any activities that are similar to practices applying the
analyses, there will be a correction for possible baseline [G]OLD-protocol. They may undertake initiatives that
differences between participants or general practices. In improve their care for older people and this may dis-
addition, we will perform sub-group analyses to investi- tort the intervention effect. Also, several factors may
gate whether certain groups of older people benefit more influence the extent to which general practices are suc-
from the [G]OLD-protocol than other groups. We will cessfully redesigned, such as the influence of existing
usethesoftwarepackageSPSSforWindows,version17.0, routines and the care providers own clinical opinion (or
forallstatisticalanalyses. “gut feelings”) on medical decision-making and referring
Stijnenetal.BMCGeriatrics2013,13:7 Page9of10
http://www.biomedcentral.com/1471-2318/13/7
older patients to adequate care and/or well-being facil- 6. ElkanR,KendrickD,DeweyM,HewittM,RobinsonJ,BlairM,WilliamsD,
ities. Although we provided the necessary guidelines and BrummellK:Effectivenessofhomebasedsupportforolderpeople:
systematicreviewandmeta-analysis.BrMedJ2001,323:719–725.
recommendations to facilitate this process, GPs and
7. StuckAE,EggerM,HammerA,MinderCE,BeckJC:Homevisitstoprevent
practice nurses may not have ignored their own medical nursinghomeadmissionandfunctionaldeclineinelderlypeople:
expertise in deciding about the diagnosis of health pro- systematicreviewandmeta-regressionanalysis.JAmMedAssoc2002,
287:1022–1029.
blems and/or referral and follow-up. The process evalu-
8. Markle-ReidM,BrowneG,WeirR,GafniA,RobertsJ,HendersonSR:The
ation will provide insight in the extent to which general effectivenessandefficiencyofhome-basednursinghealthpromotion
practices redesign their care delivery to older people forolderpeople:areviewoftheliterature.MedCareResRev2006,
63:531–569.
according to the [G]OLD-protocol. Results of the effect
9. PerryM,MokkinkHGA:Ishetzinvolomzelfstandigwonendeouderen
andprocess evaluationwillbecomeavailablein2013. systematischtebezoeken?[Isitusefultosystematicallyvisit
independentlylivingelderlypeople?]TijdschrPrakt2006,1:81–89.
10. HussA,StuckAE,RubensteinLZ,EggerM,Clough-GorrKM:Multidimensional
Abbreviations
preventivehomevisitprogramsforcommunity-dwellingolderadults:a
ADL:ActivitiesofDailyLiving;BMI:BodyMassIndex;CCM:ChronicCare
systematicreviewandmeta-analysisofrandomizedcontrolledtrials.
Model;GARS:GroningenActivityRestrictionScale;[G]OLD:GettingOLDthe
JGerontolABiolSciMedSci2008,63A:298–307.
healthyway;GP:GeneralPractitioner;IADL:InstrumentalActivitiesofDaily
11. BoumanA,VanRossumE,NelemansP,KempenGIJM,KnipschildP:
Living;n:Samplesize;PGCMoraleScale:PhiladelphiaGeriatricCenterMorale
Effectsofintensivehomevisitingprogramsforolderpeoplewithpoor
Scale;RAND-36:ResearchandDevelopment-36;SD:StandardDeviation.
healthstatus:asystematicreview.BMCHealthServRes2008,8:74.
12. RubensteinLZ,StuckAE:Preventivehomevisitsforolderpeople:defining
Competinginterests criteriaforsuccess.AgeAgeing2001,30:107–109.
Theauthorsdeclarethattheyhavenocompetinginterests. 13. EggerM:Commentary:when,where,andwhydopreventivehomevisits
work?BrMedJ2001,323:724–725.
Authors’contributions 14. VanHoutHPJ,JansenAPD,VanMarwijkHWJ,NijpelG:Preventionof
adversehealthtrajectoriesinavulnerableelderlypopulationthrough
IDasmainapplicantandMJasprojectleaderwereinvolvedinwritingthe
nursehomevisits:randomizedcontrolledtrial[ISRCTN05358495].
grantproposalforthecurrentstudy.Allauthorsparticipatedinthedesignof
JGerontolABiolSciMedSci2010,65:734–742.
thestudy.MSwasresponsiblefortherecruitmentofgeneralpracticesand
olderpeopleinthisstudy.MSdraftedthemanuscriptwithinputfromthe 15. NederlandsHuisartsenGenootschap(NHG)DutchAssociationforGeneral
otherauthors.Allauthorsread,commentedonandapprovedthefinal Practitioners:NHG-StandpuntToekomstvisieHuisartsenzorg.
manuscript. Huisartsgeneeskundevoorouderen[StatementoftheNHG,FutureofGeneral
PractitionerCare.Generalpracticemedicineforolderpeople].Utrecht:
NederlandsHuisartsenGenootschap;2007.
Acknowledgements 16. BodenheimerT,WagnerEH,GrumbachK:Improvingprimarycarefor
WewouldliketothankthelocalgeneralpractitionerorganisationsRegionale patientswithchronicillness.JAmMedAssoc2002,288:1775–1779.
HuisartsenZorgHeuvelland,HuisartsenOostelijkZuid-Limburg,Meditta,and 17. ColemanK,AustinBT,BrachC,WagnerEH:Evidenceonthechroniccare
theGPsandpracticenursesfromallparticipatinggeneralpracticesinthis modelinthenewmillennium.HealthAff(Millwood)2009,28:75–85.
studyfortheircooperation.ThisstudyisfundedbytheNetherlands 18. BoultC,KarmL,GrovesC:Improvingchroniccare:the"GuidedCare"
OrganisationforHealthResearchandDevelopment(ZonMw;grantnumber: model.PermJ2008,12:50–54.
311070303)aspartoftheDutchNationalCarefortheElderlyProgramme. 19. StalenhoefPA,HuijnenLGJ,SchonckRSM,KnottnerusJA,VanderHorstFG,
CrebolderHFJM:Huisbezoekdoordewijkverpleegkundigekanbijdragen
Authordetails aandekennisvandehuisartsoverouderepatiënten.[Homevisitsby
1DepartmentofGeneralPractice,SchoolforPublicHealthandPrimaryCare communitynursesmaycontributetogeneralpractitioner'sinformation
(CAPHRI),FacultyofHealth,MedicineandLifeSciences,MaastrichtUniversity aboutelderlypatients].NedTijdschrGeneeskd1996,140:2464–2467.
MedicalCentre,P.O.Box616,6200MDMaastricht,TheNetherlands. 20. WilsonCC,NettingFE:Comparisonofselfandhealthprofessionals'
2DepartmentofPatient&Care,MaastrichtUniversityMedicalCentre,P.O.Box ratingsofthehealthofcommunity-basedelderly.IntJAgingHumDev
5800,6202MDMaastricht,TheNetherlands.3PublicHealthService 1987,25:11–25.
South-Limburg,SchoolforPublicHealthandPrimaryCare(CAPHRI),P.O.Box 21. AlessiCA,StuckAE,AronowHU,YuhasKE,BulaCJ,MadisonR,GoldM,
2022,6160HAGeleen,TheNetherlands.4TilburgUniversity,ScientificCentre Segal-GidanF,FanelloR,RubensteinLZ,etal:Theprocessofcarein
forCareandWelfare(TRANZO),Tilburg,TheNetherlands.5NationalUniversity preventivein-homecomprehensivegeriatricassessment.JAmGeriatrSoc
ofSingapore,SawSweeHockSchoolforPublicHealth,Singapore,Singapore. 1997,45:1044–1050.
22. ChenY-M,ThompsonEA:Understandingfactorsthatinfluencesuccessof
Received:14November2012Accepted:12December2012
home-andcommunity-basedservicesinkeepingolderadultsin
Published:18January2013 communitysettings.JAgingHealth2010,22:267–291.
23. TeunissenD,VanWeelC,Lagro-JanssenT:Urinaryincontinenceinolder
References peoplelivinginthecommunity:examininghelp-seekingbehaviour.
1. BlokstraA,BaanCA,BoshuizenHC,FeenstraTL,HoogenveenRT,Picavet BrJGenPract2005,55:776–782.
HSJ,SmitHA,WijgaAH,VerschurenWMM:Vergrijzingentoekomstige 24. CummingsSR,NevittMC,KiddS:Forgettingfalls.Thelimitedaccuracyof
ziektelast.Prognosechronischeziektenprevalentie2005–2025[Impactofthe recalloffallsintheelderly.JAmGeriatrSoc1988,36:613–616.
ageingpopulationonburdenofdisease.Projectionsofchronicdisease 25. BenderP:Deceptivedistressintheelderly.AmJNurs1992,92:28–32.
prevalencefor2005–2025],RIVMrapport260401004.Bilthoven:RIVM;2007. 26. BeswickAD,ReesK,DieppeP,AyisS,Gooberman-HillR,HorwoodJ,
2. ChristensenK,DoblhammerG,RauR,VaupelJW:Ageingpopulations:the EbrahimS:Complexinterventionstoimprovephysicalfunctionand
challengesahead.Lancet2009,374:1196–1208. maintainindependentlivinginelderlypeople:asystematicreviewand
3. HébertR:Functionaldeclineinoldage.CanMedAssocJ1997,157:1037–1045. meta-analysis.Lancet2008,371:725–735.
4. StuckAE,SiuAL,WielandGD,AdamsJ,RubensteinLZ:Comprehensive 27. BeswickAD,Gooberman-HillR,SmithA,WyldeV,EbrahimS:Maintaining
geriatricassessment:ameta-analysisofcontrolledtrials.Lancet1993, independenceinolderpeople.RevClinGerontol2010,20:128–153.
342:1032–1036. 28. FagerströmL,WikbladA,NilssonJ:Anintegrativeresearchreviewof
5. VanHaastregtJCM,DiederiksJPM,VanRossumE,DeWitteLP,Crebolder preventivehomevisitsamongolderpeople–isanindividualhealth
HFJM:Effectsofpreventivehomevisitstoelderlypeoplelivinginthe resourceperspectiveavisionorareality?ScandJCaringSci2008,
community:systematicreview.BrMedJ2000,320:754–758. 23:558–568.
Stijnenetal.BMCGeriatrics2013,13:7 Page10of10
http://www.biomedcentral.com/1471-2318/13/7
29. VassM,AvlundK,HendriksenC,PhilipsonL,RiisP:Preventivehomevisits
toolderpeopleinDenmark.Why,how,bywhom,andwhen?ZGerontol
Geriatr2007,40:209–216.
30. StijnenMMN,JansenMWJ,VrijhoefHJM,Duimel-PeetersIGP:Development
ofahomevisitationprogrammefortheearlydetectionofhealth
problemsinpotentiallyfrailcommunity-dwellingolderpeopleby
generalpractices.EurJAgeing2012,doi:10.1007/s10433-012-0251-7.
31. Lemieux-CharlesL,McGuireWL:Whatdoweknowabouthealthcare
teameffectiveness?Areviewoftheliterature.MedCareResRev2006,
63:263–300.
32. RockwoodK,HowlettSE,MacKnightC,BeattieBL,BergmanH,HébertR,
HoganDB,WolfsonC,McDowellI:Prevalence,attributes,andoutcomes
offitnessandfrailtyincommunity-dwellingolderadults:Reportfrom
theCanadianStudyofHealthandAging.JGerontolABiolSciMedSci
2004,59:1310–1317.
33. NewburyJ,MarleyJ:Preventivehomevisitstoelderlypeopleinthe
community.Visitsaremostusefulforpeopleaged>75.BrMedJ2000,
321:512.
34. StuckAE,BeckJC,EggerM:Preventingdisabilityinelderlypeople.
Lancet2004,364:1641–1642.
35. VanderZeeKI,SandermanR:Hetmetenvandealgemene
gezondheidstoestandmetdeRAND-36:eenhandleiding[Measurementof
generalhealthstatuswiththeRAND-36:aguideline].Groningen:
RijksuniversiteitGroningen,NoordelijkCentrumvoor
Gezondheidsvraagstukken;1993.
36. HaysRD,SherbourneCD,MazelRM:TheRAND36-ItemHealthSurvey1.0.
HealthEcon1993,2:217–227.
37. KempenGIJM,DoeglasDM,SuurmeijerTPBM:Hetmetenvanproblemenmet
zelfredzaamheidopverzorgendenhuishoudelijkgebiedmetdeGroningen
ActiviteitenRestrictieSchaal(GARS):eenhandleiding[Measurementofdisability
inactivitiesofdailylivingandinstrumentalactivitiesofdailylivingwiththe
GroningActivityRestrictionScale(GARS):aguideline].Groningen:
RijksuniversiteitGroningen,NoordelijkCentrumvoor
Gezondheidsvraagstukken;1993.
38. KempenGIJM,MiedemaI,OrmelJ,MolenaarW:Theassessmentof
disabilitywiththeGroningenActivityRestrictionScale.Conceptual
frameworkandpsychometricproperties.SocSciMed1996,33:1601–1610.
39. HayesV,MorrisJ,WolfeC,MorganM:TheSF-36healthsurvey
questionnaire:isitsuitableforusewitholderadults?AgeAgeing1995,
24:120–125.
40. Lawton'sPGCMoraleScale;[http://www.abramsoncenter.org/pri/
documents/PGC_morale_scale.pdf]
41. SaundersRP,EvansMH,JoshiP:Developingaprocess-evaluationplanfor
assessinghealthpromotionprogramimplementation:ahow-toguide.
HealthPromotPract2005,6:134–147.
42. VanBreukelenGJP,CandelMJJM,BergerMPF:Relativeefficiencyof
unequalclustersizesforvariancecomponentestimationincluster
randomizedandmulticentretrials.StatMethodsMedRes2007,
17:439–458.
43. GreenLW,GlasgowRE:Evaluatingtherelevance,generalization,and
applicabilityofresearch.Issuesinexternalvalidationandtranslation
methodology.EvalHealthProf2006,29:126–153.
44. GlasgowRE,GreenLW,KlesgesLM,FisherE,GoldsteinMG,HaymanLL,
OckeneJK,OrleansCT:Externalvalidity:weneedtodomore.AnnBehav
Med2006,31:105–108.
45. GreenLW:Makingresearchrelevant:ifitisanevidence-basedpractice,
where'sthepractice-basedevidence?FamPract2008,25:20–24.
46. JansenM,HartingJ,EbbenN,KroonB,StappersJ,VanEngelshovenE,
Submit your next manuscript to BioMed Central
DeVriesN:Theconceptofsustainabilityandtheuseofoutcome
indicators.Acasestudytocontinueasuccessfulhealthcounselling and take full advantage of:
intervention.FamPract2008,25:32–37.
• Convenient online submission
doi:10.1186/1471-2318-13-7
• Thorough peer review
Citethisarticleas:Stijnenetal.:Earlydetectionofhealthproblemsin
potentiallyfrailcommunity-dwellingolderpeoplebygeneralpractices- • No space constraints or color figure charges
project[G]OLD:designofalongitudinal,quasi-experimentalstudy.BMC • Immediate publication on acceptance
Geriatrics201313:7.
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit